第一次baseline测试，补充psnr测定

.claude/settings.local.json

@@ -1,21 +0,0 @@
-{
-  "permissions": {
-    "allow": [
-      "Bash(conda env list:*)",
-      "Bash(mamba env:*)",
-      "Bash(micromamba env list:*)",
-      "Bash(echo:*)",
-      "Bash(git show:*)",
-      "Bash(nvidia-smi:*)",
-      "Bash(conda activate unifolm-wma)",
-      "Bash(conda info:*)",
-      "Bash(direnv allow:*)",
-      "Bash(ls:*)",
-      "Bash(for scenario in unitree_g1_pack_camera unitree_z1_dual_arm_cleanup_pencils unitree_z1_dual_arm_stackbox unitree_z1_dual_arm_stackbox_v2 unitree_z1_stackbox)",
-      "Bash(do for case in case1 case2 case3 case4)",
-      "Bash(done)",
-      "Bash(chmod:*)",
-      "Bash(ln:*)"
-    ]
-  }
-}

.envrc

@@ -1,2 +0,0 @@
-eval "$(conda shell.bash hook 2>/dev/null)"
-conda activate unifolm-wma

.gitignore

@@ -55,6 +55,7 @@ coverage.xml
 *.pot
 
 # Django stuff:
+
 local_settings.py
 db.sqlite3
 
@@ -120,15 +121,11 @@ localTest/
 fig/
 figure/
 *.mp4
-
 Data/ControlVAE.yml
 Data/Misc
 Data/Pretrained
 Data/utils.py
 Experiment/checkpoint
 Experiment/log
-
-*.ckpt
-
+ckpts/unifolm_wma_dual.ckpt
 *.0
-ckpts/unifolm_wma_dual.ckpt.prepared.pt

ckpts/LICENSE

@@ -1,439 +0,0 @@
-Attribution-NonCommercial-ShareAlike 4.0 International
-
-Copyright (c) 2016-2025 HangZhou YuShu TECHNOLOGY CO.,LTD. ("Unitree Robotics")
-
-=======================================================================
-
-Creative Commons Corporation ("Creative Commons") is not a law firm and
-does not provide legal services or legal advice. Distribution of
-Creative Commons public licenses does not create a lawyer-client or
-other relationship. Creative Commons makes its licenses and related
-information available on an "as-is" basis. Creative Commons gives no
-warranties regarding its licenses, any material licensed under their
-terms and conditions, or any related information. Creative Commons
-disclaims all liability for damages resulting from their use to the
-fullest extent possible.
-
-Using Creative Commons Public Licenses
-
-Creative Commons public licenses provide a standard set of terms and
-conditions that creators and other rights holders may use to share
-original works of authorship and other material subject to copyright
-and certain other rights specified in the public license below. The
-following considerations are for informational purposes only, are not
-exhaustive, and do not form part of our licenses.
-
-     Considerations for licensors: Our public licenses are
-     intended for use by those authorized to give the public
-     permission to use material in ways otherwise restricted by
-     copyright and certain other rights. Our licenses are
-     irrevocable. Licensors should read and understand the terms
-     and conditions of the license they choose before applying it.
-     Licensors should also secure all rights necessary before
-     applying our licenses so that the public can reuse the
-     material as expected. Licensors should clearly mark any
-     material not subject to the license. This includes other CC-
-     licensed material, or material used under an exception or
-     limitation to copyright. More considerations for licensors:
-    wiki.creativecommons.org/Considerations_for_licensors
-
-     Considerations for the public: By using one of our public
-     licenses, a licensor grants the public permission to use the
-     licensed material under specified terms and conditions. If
-     the licensor's permission is not necessary for any reason--for
-     example, because of any applicable exception or limitation to
-     copyright--then that use is not regulated by the license. Our
-     licenses grant only permissions under copyright and certain
-     other rights that a licensor has authority to grant. Use of
-     the licensed material may still be restricted for other
-     reasons, including because others have copyright or other
-     rights in the material. A licensor may make special requests,
-     such as asking that all changes be marked or described.
-     Although not required by our licenses, you are encouraged to
-     respect those requests where reasonable. More considerations
-     for the public:
-    wiki.creativecommons.org/Considerations_for_licensees
-
-=======================================================================
-
-Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
-Public License
-
-By exercising the Licensed Rights (defined below), You accept and agree
-to be bound by the terms and conditions of this Creative Commons
-Attribution-NonCommercial-ShareAlike 4.0 International Public License
-("Public License"). To the extent this Public License may be
-interpreted as a contract, You are granted the Licensed Rights in
-consideration of Your acceptance of these terms and conditions, and the
-Licensor grants You such rights in consideration of benefits the
-Licensor receives from making the Licensed Material available under
-these terms and conditions.
-
-
-Section 1 -- Definitions.
-
-  a. Adapted Material means material subject to Copyright and Similar
-     Rights that is derived from or based upon the Licensed Material
-     and in which the Licensed Material is translated, altered,
-     arranged, transformed, or otherwise modified in a manner requiring
-     permission under the Copyright and Similar Rights held by the
-     Licensor. For purposes of this Public License, where the Licensed
-     Material is a musical work, performance, or sound recording,
-     Adapted Material is always produced where the Licensed Material is
-     synched in timed relation with a moving image.
-
-  b. Adapter's License means the license You apply to Your Copyright
-     and Similar Rights in Your contributions to Adapted Material in
-     accordance with the terms and conditions of this Public License.
-
-  c. BY-NC-SA Compatible License means a license listed at
-     creativecommons.org/compatiblelicenses, approved by Creative
-     Commons as essentially the equivalent of this Public License.
-
-  d. Copyright and Similar Rights means copyright and/or similar rights
-     closely related to copyright including, without limitation,
-     performance, broadcast, sound recording, and Sui Generis Database
-     Rights, without regard to how the rights are labeled or
-     categorized. For purposes of this Public License, the rights
-     specified in Section 2(b)(1)-(2) are not Copyright and Similar
-     Rights.
-
-  e. Effective Technological Measures means those measures that, in the
-     absence of proper authority, may not be circumvented under laws
-     fulfilling obligations under Article 11 of the WIPO Copyright
-     Treaty adopted on December 20, 1996, and/or similar international
-     agreements.
-
-  f. Exceptions and Limitations means fair use, fair dealing, and/or
-     any other exception or limitation to Copyright and Similar Rights
-     that applies to Your use of the Licensed Material.
-
-  g. License Elements means the license attributes listed in the name
-     of a Creative Commons Public License. The License Elements of this
-     Public License are Attribution, NonCommercial, and ShareAlike.
-
-  h. Licensed Material means the artistic or literary work, database,
-     or other material to which the Licensor applied this Public
-     License.
-
-  i. Licensed Rights means the rights granted to You subject to the
-     terms and conditions of this Public License, which are limited to
-     all Copyright and Similar Rights that apply to Your use of the
-     Licensed Material and that the Licensor has authority to license.
-
-  j. Licensor means the individual(s) or entity(ies) granting rights
-     under this Public License.
-
-  k. NonCommercial means not primarily intended for or directed towards
-     commercial advantage or monetary compensation. For purposes of
-     this Public License, the exchange of the Licensed Material for
-     other material subject to Copyright and Similar Rights by digital
-     file-sharing or similar means is NonCommercial provided there is
-     no payment of monetary compensation in connection with the
-     exchange.
-
-  l. Share means to provide material to the public by any means or
-     process that requires permission under the Licensed Rights, such
-     as reproduction, public display, public performance, distribution,
-     dissemination, communication, or importation, and to make material
-     available to the public including in ways that members of the
-     public may access the material from a place and at a time
-     individually chosen by them.
-
-  m. Sui Generis Database Rights means rights other than copyright
-     resulting from Directive 96/9/EC of the European Parliament and of
-     the Council of 11 March 1996 on the legal protection of databases,
-     as amended and/or succeeded, as well as other essentially
-     equivalent rights anywhere in the world.
-
-  n. You means the individual or entity exercising the Licensed Rights
-     under this Public License. Your has a corresponding meaning.
-
-
-Section 2 -- Scope.
-
-  a. License grant.
-
-       1. Subject to the terms and conditions of this Public License,
-          the Licensor hereby grants You a worldwide, royalty-free,
-          non-sublicensable, non-exclusive, irrevocable license to
-          exercise the Licensed Rights in the Licensed Material to:
-
-            a. reproduce and Share the Licensed Material, in whole or
-               in part, for NonCommercial purposes only; and
-
-            b. produce, reproduce, and Share Adapted Material for
-               NonCommercial purposes only.
-
-       2. Exceptions and Limitations. For the avoidance of doubt, where
-          Exceptions and Limitations apply to Your use, this Public
-          License does not apply, and You do not need to comply with
-          its terms and conditions.
-
-       3. Term. The term of this Public License is specified in Section
-          6(a).
-
-       4. Media and formats; technical modifications allowed. The
-          Licensor authorizes You to exercise the Licensed Rights in
-          all media and formats whether now known or hereafter created,
-          and to make technical modifications necessary to do so. The
-          Licensor waives and/or agrees not to assert any right or
-          authority to forbid You from making technical modifications
-          necessary to exercise the Licensed Rights, including
-          technical modifications necessary to circumvent Effective
-          Technological Measures. For purposes of this Public License,
-          simply making modifications authorized by this Section 2(a)
-          (4) never produces Adapted Material.
-
-       5. Downstream recipients.
-
-            a. Offer from the Licensor -- Licensed Material. Every
-               recipient of the Licensed Material automatically
-               receives an offer from the Licensor to exercise the
-               Licensed Rights under the terms and conditions of this
-               Public License.
-
-            b. Additional offer from the Licensor -- Adapted Material.
-               Every recipient of Adapted Material from You
-               automatically receives an offer from the Licensor to
-               exercise the Licensed Rights in the Adapted Material
-               under the conditions of the Adapter's License You apply.
-
-            c. No downstream restrictions. You may not offer or impose
-               any additional or different terms or conditions on, or
-               apply any Effective Technological Measures to, the
-               Licensed Material if doing so restricts exercise of the
-               Licensed Rights by any recipient of the Licensed
-               Material.
-
-       6. No endorsement. Nothing in this Public License constitutes or
-          may be construed as permission to assert or imply that You
-          are, or that Your use of the Licensed Material is, connected
-          with, or sponsored, endorsed, or granted official status by,
-          the Licensor or others designated to receive attribution as
-          provided in Section 3(a)(1)(A)(i).
-
-  b. Other rights.
-
-       1. Moral rights, such as the right of integrity, are not
-          licensed under this Public License, nor are publicity,
-          privacy, and/or other similar personality rights; however, to
-          the extent possible, the Licensor waives and/or agrees not to
-          assert any such rights held by the Licensor to the limited
-          extent necessary to allow You to exercise the Licensed
-          Rights, but not otherwise.
-
-       2. Patent and trademark rights are not licensed under this
-          Public License.
-
-       3. To the extent possible, the Licensor waives any right to
-          collect royalties from You for the exercise of the Licensed
-          Rights, whether directly or through a collecting society
-          under any voluntary or waivable statutory or compulsory
-          licensing scheme. In all other cases the Licensor expressly
-          reserves any right to collect such royalties, including when
-          the Licensed Material is used other than for NonCommercial
-          purposes.
-
-
-Section 3 -- License Conditions.
-
-Your exercise of the Licensed Rights is expressly made subject to the
-following conditions.
-
-  a. Attribution.
-
-       1. If You Share the Licensed Material (including in modified
-          form), You must:
-
-            a. retain the following if it is supplied by the Licensor
-               with the Licensed Material:
-
-                 i. identification of the creator(s) of the Licensed
-                    Material and any others designated to receive
-                    attribution, in any reasonable manner requested by
-                    the Licensor (including by pseudonym if
-                    designated);
-
-                ii. a copyright notice;
-
-               iii. a notice that refers to this Public License;
-
-                iv. a notice that refers to the disclaimer of
-                    warranties;
-
-                 v. a URI or hyperlink to the Licensed Material to the
-                    extent reasonably practicable;
-
-            b. indicate if You modified the Licensed Material and
-               retain an indication of any previous modifications; and
-
-            c. indicate the Licensed Material is licensed under this
-               Public License, and include the text of, or the URI or
-               hyperlink to, this Public License.
-
-       2. You may satisfy the conditions in Section 3(a)(1) in any
-          reasonable manner based on the medium, means, and context in
-          which You Share the Licensed Material. For example, it may be
-          reasonable to satisfy the conditions by providing a URI or
-          hyperlink to a resource that includes the required
-          information.
-       3. If requested by the Licensor, You must remove any of the
-          information required by Section 3(a)(1)(A) to the extent
-          reasonably practicable.
-
-  b. ShareAlike.
-
-     In addition to the conditions in Section 3(a), if You Share
-     Adapted Material You produce, the following conditions also apply.
-
-       1. The Adapter's License You apply must be a Creative Commons
-          license with the same License Elements, this version or
-          later, or a BY-NC-SA Compatible License.
-
-       2. You must include the text of, or the URI or hyperlink to, the
-          Adapter's License You apply. You may satisfy this condition
-          in any reasonable manner based on the medium, means, and
-          context in which You Share Adapted Material.
-
-       3. You may not offer or impose any additional or different terms
-          or conditions on, or apply any Effective Technological
-          Measures to, Adapted Material that restrict exercise of the
-          rights granted under the Adapter's License You apply.
-
-
-Section 4 -- Sui Generis Database Rights.
-
-Where the Licensed Rights include Sui Generis Database Rights that
-apply to Your use of the Licensed Material:
-
-  a. for the avoidance of doubt, Section 2(a)(1) grants You the right
-     to extract, reuse, reproduce, and Share all or a substantial
-     portion of the contents of the database for NonCommercial purposes
-     only;
-
-  b. if You include all or a substantial portion of the database
-     contents in a database in which You have Sui Generis Database
-     Rights, then the database in which You have Sui Generis Database
-     Rights (but not its individual contents) is Adapted Material,
-     including for purposes of Section 3(b); and
-
-  c. You must comply with the conditions in Section 3(a) if You Share
-     all or a substantial portion of the contents of the database.
-
-For the avoidance of doubt, this Section 4 supplements and does not
-replace Your obligations under this Public License where the Licensed
-Rights include other Copyright and Similar Rights.
-
-
-Section 5 -- Disclaimer of Warranties and Limitation of Liability.
-
-  a. UNLESS OTHERWISE SEPARATELY UNDERTAKEN BY THE LICENSOR, TO THE
-     EXTENT POSSIBLE, THE LICENSOR OFFERS THE LICENSED MATERIAL AS-IS
-     AND AS-AVAILABLE, AND MAKES NO REPRESENTATIONS OR WARRANTIES OF
-     ANY KIND CONCERNING THE LICENSED MATERIAL, WHETHER EXPRESS,
-     IMPLIED, STATUTORY, OR OTHER. THIS INCLUDES, WITHOUT LIMITATION,
-     WARRANTIES OF TITLE, MERCHANTABILITY, FITNESS FOR A PARTICULAR
-     PURPOSE, NON-INFRINGEMENT, ABSENCE OF LATENT OR OTHER DEFECTS,
-     ACCURACY, OR THE PRESENCE OR ABSENCE OF ERRORS, WHETHER OR NOT
-     KNOWN OR DISCOVERABLE. WHERE DISCLAIMERS OF WARRANTIES ARE NOT
-     ALLOWED IN FULL OR IN PART, THIS DISCLAIMER MAY NOT APPLY TO YOU.
-
-  b. TO THE EXTENT POSSIBLE, IN NO EVENT WILL THE LICENSOR BE LIABLE
-     TO YOU ON ANY LEGAL THEORY (INCLUDING, WITHOUT LIMITATION,
-     NEGLIGENCE) OR OTHERWISE FOR ANY DIRECT, SPECIAL, INDIRECT,
-     INCIDENTAL, CONSEQUENTIAL, PUNITIVE, EXEMPLARY, OR OTHER LOSSES,
-     COSTS, EXPENSES, OR DAMAGES ARISING OUT OF THIS PUBLIC LICENSE OR
-     USE OF THE LICENSED MATERIAL, EVEN IF THE LICENSOR HAS BEEN
-     ADVISED OF THE POSSIBILITY OF SUCH LOSSES, COSTS, EXPENSES, OR
-     DAMAGES. WHERE A LIMITATION OF LIABILITY IS NOT ALLOWED IN FULL OR
-     IN PART, THIS LIMITATION MAY NOT APPLY TO YOU.
-
-  c. The disclaimer of warranties and limitation of liability provided
-     above shall be interpreted in a manner that, to the extent
-     possible, most closely approximates an absolute disclaimer and
-     waiver of all liability.
-
-
-Section 6 -- Term and Termination.
-
-  a. This Public License applies for the term of the Copyright and
-     Similar Rights licensed here. However, if You fail to comply with
-     this Public License, then Your rights under this Public License
-     terminate automatically.
-
-  b. Where Your right to use the Licensed Material has terminated under
-     Section 6(a), it reinstates:
-
-       1. automatically as of the date the violation is cured, provided
-          it is cured within 30 days of Your discovery of the
-          violation; or
-
-       2. upon express reinstatement by the Licensor.
-
-     For the avoidance of doubt, this Section 6(b) does not affect any
-     right the Licensor may have to seek remedies for Your violations
-     of this Public License.
-
-  c. For the avoidance of doubt, the Licensor may also offer the
-     Licensed Material under separate terms or conditions or stop
-     distributing the Licensed Material at any time; however, doing so
-     will not terminate this Public License.
-
-  d. Sections 1, 5, 6, 7, and 8 survive termination of this Public
-     License.
-
-
-Section 7 -- Other Terms and Conditions.
-
-  a. The Licensor shall not be bound by any additional or different
-     terms or conditions communicated by You unless expressly agreed.
-
-  b. Any arrangements, understandings, or agreements regarding the
-     Licensed Material not stated herein are separate from and
-     independent of the terms and conditions of this Public License.
-
-
-Section 8 -- Interpretation.
-
-  a. For the avoidance of doubt, this Public License does not, and
-     shall not be interpreted to, reduce, limit, restrict, or impose
-     conditions on any use of the Licensed Material that could lawfully
-     be made without permission under this Public License.
-
-  b. To the extent possible, if any provision of this Public License is
-     deemed unenforceable, it shall be automatically reformed to the
-     minimum extent necessary to make it enforceable. If the provision
-     cannot be reformed, it shall be severed from this Public License
-     without affecting the enforceability of the remaining terms and
-     conditions.
-
-  c. No term or condition of this Public License will be waived and no
-     failure to comply consented to unless expressly agreed to by the
-     Licensor.
-
-  d. Nothing in this Public License constitutes or may be interpreted
-     as a limitation upon, or waiver of, any privileges and immunities
-     that apply to the Licensor or You, including from the legal
-     processes of any jurisdiction or authority.
-
-=======================================================================
-
-Creative Commons is not a party to its public
-licenses. Notwithstanding, Creative Commons may elect to apply one of
-its public licenses to material it publishes and in those instances
-will be considered the “Licensor.” The text of the Creative Commons
-public licenses is dedicated to the public domain under the CC0 Public
-Domain Dedication. Except for the limited purpose of indicating that
-material is shared under a Creative Commons public license or as
-otherwise permitted by the Creative Commons policies published at
-creativecommons.org/policies, Creative Commons does not authorize the
-use of the trademark "Creative Commons" or any other trademark or logo
-of Creative Commons without its prior written consent including,
-without limitation, in connection with any unauthorized modifications
-to any of its public licenses or any other arrangements,
-understandings, or agreements concerning use of licensed material. For
-the avoidance of doubt, this paragraph does not form part of the
-public licenses.
-
-Creative Commons may be contacted at creativecommons.org.

ckpts/README.md

@@ -1,38 +0,0 @@
----
-tags:
-- robotics
----
-
-# UnifoLM-WMA-0: A World-Model-Action (WMA) Framework under UnifoLM Family
-<p style="font-size: 1.2em;">
-    <a href="https://unigen-x.github.io/unifolm-world-model-action.github.io"><strong>Project Page</strong></a> | 
-    <a href="https://github.com/unitreerobotics/unifolm-world-model-action"><strong>Code</strong></a> |
-    <a href="https://huggingface.co/unitreerobotics/datasets"><strong>Dataset</strong></a> 
-  </p>
-<div align="center">
-  <div align="justify">
-    <b>UnifoLM-WMA-0</b> is Unitree‘s first open-source world-model–action architecture spanning multiple types of robotic embodiments, designed specifically for general-purpose robot learning. Its core component is a world-model capable of understanding the physical interactions between robots and the environments. This world-model provides two key functions: (a) <b>Simulation Engine</b> – operates as an interactive simulator to generate synthetic data for robot learning; (b) <b>Policy Enhancement</b> – connects with an action head and, by predicting future interaction processes with the world-model, further optimizes decision-making performance.
-  </div>
-</div>
-
-## 🦾 Real Robot Deployment
-| <img src="assets/real_z1_stackbox.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> | <img src="assets/real_dual_stackbox.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> |
-|:---:|:---:|
-| <img src="assets/real_cleanup_pencils.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> | <img src="assets/real_g1_pack_camera.gif" style="border:none;box-shadow:none;margin:0;padding:0;" /> |
-
-**Note: the top-right window shows the world model’s prediction of future environmental changes.**
-
-## License
-The model is released under the CC BY-NC-SA 4.0 license as found in the [LICENSE](https://huggingface.co/unitreerobotics/UnifoLM-WMA-0/blob/main/LICENSE). You are responsible for ensuring that your use of Unitree AI Models complies with all applicable laws.
-
-## Model Architecture
-![Demo](assets/world_model_interaction.gif)
-
-## Citation
-```
-@misc{unifolm-wma-0,
-  author = {Unitree},
-  title  = {UnifoLM-WMA-0: A World-Model-Action (WMA) Framework under UnifoLM Family},
-  year   = {2025},
-}
-```

docs/architecture_overview.typ

@@ -1,122 +0,0 @@
-== Task Comprehension: Diffusion Model and UnifoLM-WMA
-
-This section provides a comprehensive overview of the UnifoLM-WMA-0 deep learning architecture, serving as a practical foundation for the optimization strategies discussed in subsequent sections.
-
-=== Overall Inference Pipeline
-
-UnifoLM-WMA-0 is Unitree Robotics' open-source World-Model-Action framework. Its core task is to predict future video frame sequences along with the corresponding robot action and state trajectories, given a current observation image and a text instruction. The model operates in an interactive simulation mode: each iteration consumes the previous prediction as input and generates the next segment of video and actions, thereby forming a closed-loop rollout. A single iteration of this pipeline can be decomposed into four sequential stages — condition encoding, VAE encoding, DDIM diffusion sampling, and VAE decoding — each of which is described below.
-
-==== Condition Encoding
-
-The condition encoding stage transforms raw multi-modal inputs into a unified context vector that guides the diffusion denoising process, through three parallel encoding paths. On the image side, the input observation image (320#sym.times 512) is processed by a frozen OpenCLIP ViT-H-14 vision encoder, then compressed through a Resampler — a Perceiver-based cross-attention module (4 layers, 12 heads, dim\_head=64, embed\_dim 1280 #sym.arrow 1024) — into 16 image condition tokens per frame, yielding $16 times T = 256$ image tokens for T=16 frames.
-
-On the text side, the instruction is encoded by a frozen OpenCLIP text encoder (`FrozenOpenCLIPEmbedder`, penultimate layer output) into 77 tokens of dimension 1024, computed once and reused across all DDIM steps. On the state side, the robot proprioceptive state (dim 16) is mapped through a SATokenProjector (Perceiver Attention, 1 layer, 16 heads, dim\_head=64, 16 learnable queries) into 16 tokens of dimension 1024.
-
-These three token sets are concatenated to form the unified context vector: `[agent_state(2) | agent_action(16) | text(77) | image(256)]`, totaling 351 tokens per cross-attention operation.
-
-==== VAE Encoding
-
-The observation images are encoded into a compact latent space through an AutoencoderKL (`autoencoder.py`) — a variational autoencoder regularized by KL divergence. The encoder follows a convolutional architecture with 4-level channel multipliers [1, 2, 4, 4] (base channels ch=128, yielding channel widths [128, 256, 512, 512]), 2 residual blocks per level, and a latent channel count of z\_channels=4. The input RGB frames at resolution 320#sym.times 512 are encoded into latent representations at 1/8 spatial resolution, producing tensors of shape `(B, 4, T, 40, 64)`.
-
-A critical configuration parameter is `perframe_ae=True`, which means the VAE processes each of the T=16 frames independently rather than as a 3D volume. While this per-frame strategy avoids the memory overhead of volumetric convolutions, it introduces a sequential loop of T forward passes through the encoder — a point worth noting for latency optimization. The latent representations are scaled by a fixed factor of `scale_factor=0.18215` before being fed into the diffusion process.
-
-==== DDIM Diffusion Sampling
-
-This is the core time-consuming part of inference. A DDIM (Denoising Diffusion Implicit Models) sampler (`ddim.py`) is employed with a default of 50 denoising steps. The diffusion process is parameterized with v-prediction (`parameterization="v"`), 1000 training timesteps, and a linear beta schedule from `linear_start=0.00085` to `linear_end=0.012`, with zero-SNR terminal rescaling enabled (`rescale_betas_zero_snr=True`) and dynamic rescaling applied at `base_scale=0.7` to stabilize generation quality.
-
-Unlike standard video diffusion models that only predict denoised video latents, UnifoLM-WMA simultaneously produces three outputs per step: a video latent prediction `y` of shape `(B, 4, T, 40, 64)`, an action trajectory prediction `a_y` of shape `(B, T, 16)`, and a state trajectory prediction `s_y` of shape `(B, T, 16)`. The three predictions share the same diffusion timestep but employ heterogeneous noise schedules — the video stream uses the DDPM schedule with v-prediction, while the action and state streams use a `DDIMScheduler` from the `diffusers` library with epsilon-prediction and a `squaredcos_cap_v2` beta schedule. This design allows each modality to adopt its optimal denoising strategy.
-
-The sampler also supports classifier-free guidance with `unconditional_guidance_scale` and guidance rescaling, applied only to the video stream to balance generation quality and diversity.
-
-==== VAE Decoding
-
-After the DDIM sampling loop completes, the denoised video latent tensor $x_0$ of shape `(B, 4, T, 40, 64)` is decoded back to RGB pixel space through the AutoencoderKL decoder. Due to the `perframe_ae=True` configuration, decoding is likewise performed frame-by-frame: each of the T=16 latent frames is individually inverse-scaled by $1 slash "scale_factor"$, passed through the decoder's convolutional transpose layers, and reconstructed to a 320#sym.times 512 RGB frame.
-
-In the interactive simulation mode, the decoded video serves a dual purpose — providing the observation image for the next iteration's condition encoding (only the first `exe_steps` frames are needed) and producing the final output video for visualization and evaluation. The action and state trajectories predicted by the DDIM loop are directly used for robot control without further decoding.
-
-=== WMAModel Backbone: Dual-UNet Collaborative Architecture
-
-The WMAModel (`wma_model.py:326`) is the core neural network invoked at every DDIM step, employing a unique dual-UNet collaborative architecture that jointly predicts video, actions, and states within a single forward pass. This tightly-coupled design enables the action and state predictions to directly leverage the rich spatiotemporal features extracted by the video generation backbone, rather than treating them as independent prediction heads.
-
-==== Video UNet
-
-The primary backbone is a 2D convolution-based UNet with temporal extensions. Its key configuration is summarized in the following table:
-
-#figure(
-  table(
-    columns: (3fr, 5fr),
-    [*Parameter*], [*Value*],
-    [Input / Output channels], [8 (4 latent + 4 conditioning) / 4],
-    [Base model channels], [320],
-    [Channel multipliers], [\[1, 2, 4, 4\] #sym.arrow widths \[320, 640, 1280, 1280\]],
-    [Residual blocks per level], [2],
-    [Attention resolutions], [\[4, 2, 1\] (3 of 4 resolution levels)],
-    [Attention head channels], [64],
-    [Transformer depth], [1 per attention resolution],
-    [Context dimension], [1024],
-    [Temporal length], [16 frames],
-  ),
-  caption: [Video UNet configuration parameters.],
-)
-
-The UNet follows the classic encoder-middle-decoder structure with skip connections. At each attention-enabled resolution level, every ResBlock is followed by two transformer modules: a SpatialTransformer that performs spatial self-attention among all $H times W$ tokens within each frame followed by cross-attention with the 351-token context vector, and a TemporalTransformer that performs self-attention among T=16 time-step tokens at each spatial position (configured with `temporal_selfatt_only=True`, i.e., no cross-attention).
-
-During the forward pass, intermediate feature maps are collected after each Downsample layer and the middle block, reshaped from $(B times T, C, H, W)$ to $(B, T, C, H, W)$, accumulating 10 multi-scale feature maps in `hs_a` — the bridge to the Action/State UNets.
-
-==== Action UNet and State UNet
-
-The Action UNet (`conditional_unet1d.py`) is a 1D convolutional UNet specifically designed for predicting robot action trajectories. Its configuration is as follows:
-
-#figure(
-  table(
-    columns: (3fr, 5fr),
-    [*Parameter*], [*Value*],
-    [Input dimension], [16 (agent\_action\_dim)],
-    [Down channel widths], [\[256, 512, 1024, 2048\]],
-    [Kernel size], [5],
-    [GroupNorm groups], [8],
-    [Diffusion step embedding dim], [128],
-    [Horizon], [16],
-    [Action projection dim], [32],
-  ),
-  caption: [Action UNet (ConditionalUnet1D) configuration parameters.],
-)
-
-The Action UNet receives the 10 `hs_a` feature maps from the Video UNet as visual conditioning. The conditioning pipeline involves three stages: (1) SpatialSoftmax compresses each 2D feature map into keypoint coordinates $(B times T, C, 2)$; (2) the compressed features are concatenated with the diffusion timestep embedding and observation encoding (ResNet-18 `MultiImageObsEncoder`), then injected via FiLM modulation to produce per-channel scale/bias for the 1D convolution blocks; (3) `ActionLatentImageCrossAttention` enables action tokens to cross-attend to the Video UNet's spatiotemporal features, allowing visually-grounded action planning.
-
-The input action tensor $(B, T, 16)$ is projected to act\_proj\_dim=32, processed through the 1D UNet, then projected back to $(B, T, 16)$.
-
-The State UNet is an identical `ConditionalUnet1D` instance with the same hyperparameters, operating on the state tensor `x_state` $(B, T, 16)$ instead of the action tensor.
-
-A critical optimization observation: the Action and State UNets are computationally independent — sharing read-only inputs with no data dependencies. The original code executes them sequentially, leaving significant room for CUDA stream parallelization.
-
-=== Multi-Level Design of Attention Mechanisms
-
-The attention mechanisms in UnifoLM-WMA constitute the core computational bottleneck of inference. Their design encompasses four distinct levels, each serving a different purpose in the model's spatiotemporal reasoning, and understanding their structure is essential for identifying optimization opportunities.
-
-The first level is *spatial self-attention* within the SpatialTransformer. For a latent frame at resolution $H times W$, the token count is $H times W$ (e.g., $40 times 64 = 2560$ at the highest resolution). Implemented via xformers `memory_efficient_attention`, reducing peak memory from $O(N^2)$ to $O(N)$. Q/K/V use bias-free linear layers with head count = channel\_dim / num\_head\_channels (e.g., 1280/64 = 20 heads).
-
-The second level is *multi-source cross-attention*, the most distinctive design in UnifoLM-WMA. The unified context vector is split into four semantic sources, each with dedicated K/V projection layers:
-
-#figure(
-  table(
-    columns: (2fr, 1fr, 3fr, 2fr),
-    [*Source*], [*Tokens*], [*K/V Projections*], [*Scale*],
-    [Text], [77], [`to_k` / `to_v` (shared base)], [1.0],
-    [Image], [16#sym.times T], [`to_k_ip` / `to_v_ip`], [`image_cross_attention_scale`],
-    [Agent state], [2], [`to_k_as` / `to_v_as`], [`agent_state_cross_attention_scale`],
-    [Agent action], [16], [`to_k_aa` / `to_v_aa`], [`agent_action_cross_attention_scale`],
-  ),
-  caption: [Multi-source cross-attention configuration.],
-)
-
-The Query vector Q is always derived from the video latent features via `to_q`. For each of the four sources, independent attention scores are computed — $"softmax"(Q dot K_i^T \/ sqrt(d)) dot V_i$ — producing four separate attention outputs. These outputs are then combined via weighted summation:
-
-$ "out" = "out"_"text" + alpha_"img" dot "out"_"ip" + alpha_"state" dot "out"_"as" + alpha_"action" dot "out"_"aa" $
-
-In the current configuration, `cross_attention_scale_learnable=False` (fixed scales). This decoupled design adds 8 extra linear layers versus standard single-source cross-attention, creating opportunities for KV fusion optimization.
-
-The third level is *temporal self-attention* within the TemporalTransformer. The input $(B, C, T, H, W)$ is reshaped to $(B times H times W, C, T)$, so each spatial position becomes an independent batch element and T=16 time steps form the token sequence. Supports relative position encoding via a `RelativePosition` module and optional causal masks; current configuration uses bidirectional temporal attention.
-
-The fourth level is *action-latent-image cross-attention* in the `ActionLatentImageCrossAttention` module. Action tokens $(B, "action_dim", "act_proj_dim")$ as Query cross-attend to Video UNet features reshaped to $(B, T times H times W, C)$ as Key/Value. A `BasicTransformerBlock` (depth=1) performs action self-attention then cross-attention to video features, with zero-initialized `proj_out` and residual connection. This mechanism is the key bridge enabling the action head to access the visual world model's internal representations.
-

pyproject.toml

@@ -19,13 +19,13 @@ dependencies = [
     "pytorch-lightning==1.9.3",
     "pyyaml==6.0",
     "setuptools==65.6.3",
-    #"torch==2.3.1",
-    #"torchvision==0.18.1",
+    "torch==2.3.1",
+    "torchvision==0.18.1",
     "tqdm==4.66.5",
     "transformers==4.40.1",
     "moviepy==1.0.3",
     "av==12.3.0",
-    #"xformers==0.0.27",
+    "xformers==0.0.27",
     "gradio==4.39.0",
     "timm==0.9.10",
     "scikit-learn==1.5.1",

run_all_cases_20260211_173422.log

@@ -1,37 +0,0 @@
-2026-02-11 17:34:29.188470: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-11 17:34:29.238296: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-11 17:34:29.238342: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-11 17:34:29.239649: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-11 17:34:29.247152: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
-To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-11 17:34:30.172640: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
-Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
-INFO:root:***** Configing Data *****
->>> unitree_z1_stackbox: 1 data samples loaded.
->>> unitree_z1_stackbox: data stats loaded.
->>> unitree_z1_stackbox: normalizer initiated.
->>> unitree_z1_dual_arm_stackbox: 1 data samples loaded.
->>> unitree_z1_dual_arm_stackbox: data stats loaded.
->>> unitree_z1_dual_arm_stackbox: normalizer initiated.
->>> unitree_z1_dual_arm_stackbox_v2: 1 data samples loaded.
->>> unitree_z1_dual_arm_stackbox_v2: data stats loaded.
->>> unitree_z1_dual_arm_stackbox_v2: normalizer initiated.
->>> unitree_z1_dual_arm_cleanup_pencils: 1 data samples loaded.
->>> unitree_z1_dual_arm_cleanup_pencils: data stats loaded.
->>> unitree_z1_dual_arm_cleanup_pencils: normalizer initiated.
->>> unitree_g1_pack_camera: 1 data samples loaded.
->>> unitree_g1_pack_camera: data stats loaded.
->>> unitree_g1_pack_camera: normalizer initiated.
->>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
->>> Generate 16 frames under each generation ...
-DEBUG:h5py._conv:Creating converter from 3 to 5
-DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
-DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
-DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-

scripts/evaluation/base_model_inference.py

@@ -16,9 +16,6 @@ from collections import OrderedDict
 from unifolm_wma.models.samplers.ddim import DDIMSampler
 from unifolm_wma.utils.utils import instantiate_from_config
 
-torch.backends.cuda.matmul.allow_tf32 = True
-torch.backends.cudnn.allow_tf32 = True
-
 
 def get_filelist(data_dir: str, postfixes: list[str]) -> list[str]:
     """

scripts/evaluation/real_eval_server.py

@@ -19,9 +19,6 @@ from fastapi.responses import JSONResponse
 from typing import Any, Dict, Optional, Tuple, List
 from datetime import datetime
 
-torch.backends.cuda.matmul.allow_tf32 = True
-torch.backends.cudnn.allow_tf32 = True
-
 from unifolm_wma.utils.utils import instantiate_from_config
 from unifolm_wma.models.samplers.ddim import DDIMSampler
 

scripts/evaluation/world_model_interaction.py

@@ -9,9 +9,6 @@ import logging
 import einops
 import warnings
 import imageio
-import atexit
-import multiprocessing as mp
-from concurrent.futures import ThreadPoolExecutor
 
 from pytorch_lightning import seed_everything
 from omegaconf import OmegaConf
@@ -19,12 +16,8 @@ from tqdm import tqdm
 from einops import rearrange, repeat
 from collections import OrderedDict
 from torch import nn
-from eval_utils import populate_queues
+from eval_utils import populate_queues, log_to_tensorboard
 from collections import deque
-from typing import Optional, List, Any
-
-torch.backends.cuda.matmul.allow_tf32 = True
-torch.backends.cudnn.allow_tf32 = True
 from torch import Tensor
 from torch.utils.tensorboard import SummaryWriter
 from PIL import Image
@@ -157,107 +150,6 @@ def save_results(video: Tensor, filename: str, fps: int = 8) -> None:
                                options={'crf': '10'})
 
 
-# ========== Async I/O ==========
-_io_executor: Optional[ThreadPoolExecutor] = None
-_io_futures: List[Any] = []
-
-
-def _get_io_executor() -> ThreadPoolExecutor:
-    global _io_executor
-    if _io_executor is None:
-        _io_executor = ThreadPoolExecutor(max_workers=2)
-    return _io_executor
-
-
-def _flush_io():
-    """Wait for all pending async I/O to finish."""
-    global _io_futures
-    for fut in _io_futures:
-        try:
-            fut.result()
-        except Exception as e:
-            print(f">>> [async I/O] error: {e}")
-    _io_futures.clear()
-
-
-atexit.register(_flush_io)
-
-
-def _save_results_sync(video_cpu: Tensor, filename: str, fps: int) -> None:
-    """Synchronous save on CPU tensor (runs in background thread)."""
-    video = torch.clamp(video_cpu.float(), -1., 1.)
-    n = video.shape[0]
-    video = video.permute(2, 0, 1, 3, 4)
-    frame_grids = [
-        torchvision.utils.make_grid(framesheet, nrow=int(n), padding=0)
-        for framesheet in video
-    ]
-    grid = torch.stack(frame_grids, dim=0)
-    grid = (grid + 1.0) / 2.0
-    grid = (grid * 255).to(torch.uint8).permute(0, 2, 3, 1)
-    torchvision.io.write_video(filename,
-                               grid,
-                               fps=fps,
-                               video_codec='h264',
-                               options={'crf': '10'})
-
-
-def save_results_async(video: Tensor, filename: str, fps: int = 8) -> None:
-    """Submit video saving to background thread pool."""
-    video_cpu = video.detach().cpu()
-    fut = _get_io_executor().submit(_save_results_sync, video_cpu, filename, fps)
-    _io_futures.append(fut)
-
-
-def _log_to_tb_sync(writer, video_cpu: Tensor, tag: str, fps: int) -> None:
-    """Synchronous TensorBoard log on CPU tensor (runs in background thread)."""
-    if video_cpu.dim() == 5:
-        n = video_cpu.shape[0]
-        video = video_cpu.permute(2, 0, 1, 3, 4)
-        frame_grids = [
-            torchvision.utils.make_grid(framesheet, nrow=int(n), padding=0)
-            for framesheet in video
-        ]
-        grid = torch.stack(frame_grids, dim=0)
-        grid = (grid + 1.0) / 2.0
-        grid = grid.unsqueeze(dim=0)
-        writer.add_video(tag, grid, fps=fps)
-
-
-def log_to_tensorboard_async(writer, data: Tensor, tag: str, fps: int = 10) -> None:
-    """Submit TensorBoard logging to background thread pool."""
-    if isinstance(data, torch.Tensor) and data.dim() == 5:
-        data_cpu = data.detach().cpu()
-        fut = _get_io_executor().submit(_log_to_tb_sync, writer, data_cpu, tag, fps)
-        _io_futures.append(fut)
-
-
-def _video_tensor_to_frames(video: Tensor) -> np.ndarray:
-    video = torch.clamp(video.float(), -1., 1.)
-    n = video.shape[0]
-    video = video.permute(2, 0, 1, 3, 4)
-    frame_grids = [
-        torchvision.utils.make_grid(f, nrow=int(n), padding=0) for f in video
-    ]
-    grid = torch.stack(frame_grids, dim=0)
-    grid = ((grid + 1.0) / 2.0 * 255).to(torch.uint8).permute(0, 2, 3, 1)
-    return grid.numpy()[:, :, :, ::-1]
-
-
-def _video_writer_process(q: mp.Queue, filename: str, fps: int):
-    frames = []
-    while True:
-        item = q.get()
-        if item is None:
-            break
-        frames.append(_video_tensor_to_frames(item))
-    if frames:
-        grid = np.concatenate(frames, axis=0)
-        grid = torch.from_numpy(grid[:, :, :, ::-1].copy())  # BGR → RGB
-        torchvision.io.write_video(filename, grid, fps=fps,
-                                   video_codec='h264', options={'crf': '10'})
-
-
 def get_init_frame_path(data_dir: str, sample: dict) -> str:
     """Construct the init_frame path from directory and sample metadata.
 
@@ -435,8 +327,7 @@ def image_guided_synthesis_sim_mode(
         timestep_spacing: str = 'uniform',
         guidance_rescale: float = 0.0,
         sim_mode: bool = True,
-        decode_video: bool = True,
-        **kwargs) -> tuple[torch.Tensor | None, torch.Tensor, torch.Tensor]:
+        **kwargs) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
     """
     Performs image-guided video generation in a simulation-style mode with optional multimodal guidance (image, state, action, text).
 
@@ -459,13 +350,10 @@ def image_guided_synthesis_sim_mode(
         timestep_spacing (str): Timestep sampling method in DDIM sampler. Typically "uniform" or "linspace".
         guidance_rescale (float): Guidance rescaling factor to mitigate overexposure from classifier-free guidance.
         sim_mode (bool): Whether to perform world-model interaction or decision-making using the world-model.
-        decode_video (bool): Whether to decode latent samples to pixel-space video.
-            Set to False to skip VAE decode for speed when only actions/states are needed.
         **kwargs: Additional arguments passed to the DDIM sampler.
 
     Returns:
-        batch_variants (torch.Tensor | None): Predicted pixel-space video frames [B, C, T, H, W],
-            or None when decode_video=False.
+        batch_variants (torch.Tensor): Predicted pixel-space video frames [B, C, T, H, W].
         actions (torch.Tensor): Predicted action sequences [B, T, D] from diffusion decoding.
         states (torch.Tensor): Predicted state sequences [B, T, D] from diffusion decoding.
     """
@@ -477,9 +365,8 @@ def image_guided_synthesis_sim_mode(
 
     img = observation['observation.images.top'].permute(0, 2, 1, 3, 4)
     cond_img = rearrange(img, 'b o c h w -> (b o) c h w')[-1:]
-    with torch.cuda.amp.autocast(dtype=torch.float16):
-        cond_img_emb = model.embedder(cond_img)
-        cond_img_emb = model.image_proj_model(cond_img_emb)
+    cond_img_emb = model.embedder(cond_img)
+    cond_img_emb = model.image_proj_model(cond_img_emb)
 
     if model.model.conditioning_key == 'hybrid':
         z = get_latent_z(model, img.permute(0, 2, 1, 3, 4))
@@ -493,12 +380,11 @@ def image_guided_synthesis_sim_mode(
         prompts = [""] * batch_size
     cond_ins_emb = model.get_learned_conditioning(prompts)
 
-    with torch.cuda.amp.autocast(dtype=torch.float16):
-        cond_state_emb = model.state_projector(observation['observation.state'])
-        cond_state_emb = cond_state_emb + model.agent_state_pos_emb
+    cond_state_emb = model.state_projector(observation['observation.state'])
+    cond_state_emb = cond_state_emb + model.agent_state_pos_emb
 
-        cond_action_emb = model.action_projector(observation['action'])
-        cond_action_emb = cond_action_emb + model.agent_action_pos_emb
+    cond_action_emb = model.action_projector(observation['action'])
+    cond_action_emb = cond_action_emb + model.agent_action_pos_emb
 
     if not sim_mode:
         cond_action_emb = torch.zeros_like(cond_action_emb)
@@ -520,8 +406,6 @@ def image_guided_synthesis_sim_mode(
     kwargs.update({"unconditional_conditioning_img_nonetext": None})
     cond_mask = None
     cond_z0 = None
-    batch_variants = None
-    samples = None
     if ddim_sampler is not None:
         samples, actions, states, intermedia = ddim_sampler.sample(
             S=ddim_steps,
@@ -540,12 +424,11 @@ def image_guided_synthesis_sim_mode(
             guidance_rescale=guidance_rescale,
             **kwargs)
 
-        if decode_video:
-            # Reconstruct from latent to pixel space
-            batch_images = model.decode_first_stage(samples)
-            batch_variants = batch_images
+        # Reconstruct from latent to pixel space
+        batch_images = model.decode_first_stage(samples)
+        batch_variants = batch_images
 
-    return batch_variants, actions, states, samples
+    return batch_variants, actions, states
 
 
 def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
@@ -570,67 +453,26 @@ def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
     csv_path = os.path.join(args.prompt_dir, f"{args.dataset}.csv")
     df = pd.read_csv(csv_path)
 
-    # Load config (always needed for data setup)
+    # Load config
     config = OmegaConf.load(args.config)
+    config['model']['params']['wma_config']['params'][
+        'use_checkpoint'] = False
+    model = instantiate_from_config(config.model)
+    model.perframe_ae = args.perframe_ae
+    assert os.path.exists(args.ckpt_path), "Error: checkpoint Not Found!"
+    model = load_model_checkpoint(model, args.ckpt_path)
+    model.eval()
+    print(f'>>> Load pre-trained model ...')
 
-    prepared_path = args.ckpt_path + ".prepared.pt"
-    if os.path.exists(prepared_path):
-        # ---- Fast path: load the fully-prepared model ----
-        print(f">>> Loading prepared model from {prepared_path} ...")
-        model = torch.load(prepared_path,
-                           map_location=f"cuda:{gpu_no}",
-                           weights_only=False,
-                           mmap=True)
-        model.eval()
-        print(f">>> Prepared model loaded.")
-    else:
-        # ---- Normal path: construct + load checkpoint ----
-        config['model']['params']['wma_config']['params'][
-            'use_checkpoint'] = False
-        model = instantiate_from_config(config.model)
-        model.perframe_ae = args.perframe_ae
-
-        assert os.path.exists(args.ckpt_path), "Error: checkpoint Not Found!"
-        model = load_model_checkpoint(model, args.ckpt_path)
-        model.eval()
-        model = model.cuda(gpu_no)
-        print(f'>>> Load pre-trained model ...')
-
-        # Save prepared model for fast loading next time
-        print(f">>> Saving prepared model to {prepared_path} ...")
-        torch.save(model, prepared_path)
-        print(f">>> Prepared model saved ({os.path.getsize(prepared_path) / 1024**3:.1f} GB).")
-
-    # ---- FP16: convert diffusion backbone + conditioning modules ----
-    model.model.to(torch.float16)
-    model.model.diffusion_model.dtype = torch.float16
-    print(">>> Diffusion backbone (model.model) converted to FP16.")
-
-    # Projectors / MLP → FP16
-    model.image_proj_model.half()
-    model.state_projector.half()
-    model.action_projector.half()
-    print(">>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.")
-
-    # Text/image encoders → FP16
-    model.cond_stage_model.half()
-    model.embedder.half()
-    print(">>> Encoders (cond_stage_model, embedder) converted to FP16.")
-
-    # Build normalizer (always needed, independent of model loading path)
+    # Build unnomalizer
     logging.info("***** Configing Data *****")
     data = instantiate_from_config(config.data)
     data.setup()
     print(">>> Dataset is successfully loaded ...")
 
+    model = model.cuda(gpu_no)
     device = get_device_from_parameters(model)
 
-    # Fuse KV projections in attention layers (to_k + to_v → to_kv)
-    from unifolm_wma.modules.attention import CrossAttention
-    kv_count = sum(1 for m in model.modules()
-                   if isinstance(m, CrossAttention) and m.fuse_kv())
-    print(f"    ✓ KV fused: {kv_count} attention layers")
-
     # Run over data
     assert (args.height % 16 == 0) and (
         args.width % 16
@@ -675,13 +517,8 @@ def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
             # For saving environmental changes in world-model
             sample_save_dir = f'{video_save_dir}/wm/{fs}'
             os.makedirs(sample_save_dir, exist_ok=True)
-            # Writer process for incremental video saving
-            sample_full_video_file = f"{video_save_dir}/../{sample['videoid']}_full_fs{fs}.mp4"
-            write_q = mp.Queue()
-            writer_proc = mp.Process(
-                target=_video_writer_process,
-                args=(write_q, sample_full_video_file, args.save_fps))
-            writer_proc.start()
+            # For collecting interaction videos
+            wm_video = []
             # Initialize observation queues
             cond_obs_queues = {
                 "observation.images.top":
@@ -737,7 +574,7 @@ def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
 
                 # Use world-model in policy to generate action
                 print(f'>>> Step {itr}: generating actions ...')
-                pred_videos_0, pred_actions, _, _ = image_guided_synthesis_sim_mode(
+                pred_videos_0, pred_actions, _ = image_guided_synthesis_sim_mode(
                     model,
                     sample['instruction'],
                     observation,
@@ -750,8 +587,7 @@ def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
                     fs=model_input_fs,
                     timestep_spacing=args.timestep_spacing,
                     guidance_rescale=args.guidance_rescale,
-                    sim_mode=False,
-                    decode_video=not args.fast_policy_no_decode)
+                    sim_mode=False)
 
                 # Update future actions in the observation queues
                 for idx in range(len(pred_actions[0])):
@@ -779,7 +615,7 @@ def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
 
                 # Interaction with the world-model
                 print(f'>>> Step {itr}: interacting with world model ...')
-                pred_videos_1, _, pred_states, wm_samples = image_guided_synthesis_sim_mode(
+                pred_videos_1, _, pred_states = image_guided_synthesis_sim_mode(
                     model,
                     "",
                     observation,
@@ -792,16 +628,12 @@ def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
                     fs=model_input_fs,
                     text_input=False,
                     timestep_spacing=args.timestep_spacing,
-                    guidance_rescale=args.guidance_rescale,
-                    decode_video=False)
-
-                # Decode only the last frame for CLIP embedding in next iteration
-                last_frame_pixel = model.decode_first_stage(wm_samples[:, :, -1:, :, :])
+                    guidance_rescale=args.guidance_rescale)
 
                 for idx in range(args.exe_steps):
                     observation = {
                         'observation.images.top':
-                        last_frame_pixel[0, :, 0:1].permute(1, 0, 2, 3),
+                        pred_videos_1[0][:, idx:idx + 1].permute(1, 0, 2, 3),
                         'observation.state':
                         torch.zeros_like(pred_states[0][idx:idx + 1]) if
                         args.zero_pred_state else pred_states[0][idx:idx + 1],
@@ -812,26 +644,42 @@ def run_inference(args: argparse.Namespace, gpu_num: int, gpu_no: int) -> None:
                     cond_obs_queues = populate_queues(cond_obs_queues,
                                                       observation)
 
-                # Save the imagen videos for decision-making (async)
-                if pred_videos_0 is not None:
-                    sample_tag = f"{args.dataset}-vid{sample['videoid']}-dm-fs-{fs}/itr-{itr}"
-                    log_to_tensorboard_async(writer,
-                                             pred_videos_0,
-                                             sample_tag,
-                                             fps=args.save_fps)
+                # Save the imagen videos for decision-making
+                sample_tag = f"{args.dataset}-vid{sample['videoid']}-dm-fs-{fs}/itr-{itr}"
+                log_to_tensorboard(writer,
+                                   pred_videos_0,
+                                   sample_tag,
+                                   fps=args.save_fps)
+                # Save videos environment changes via world-model interaction
+                sample_tag = f"{args.dataset}-vid{sample['videoid']}-wd-fs-{fs}/itr-{itr}"
+                log_to_tensorboard(writer,
+                                   pred_videos_1,
+                                   sample_tag,
+                                   fps=args.save_fps)
+
+                # Save the imagen videos for decision-making
+                sample_video_file = f'{video_save_dir}/dm/{fs}/itr-{itr}.mp4'
+                save_results(pred_videos_0.cpu(),
+                             sample_video_file,
+                             fps=args.save_fps)
+                # Save videos environment changes via world-model interaction
+                sample_video_file = f'{video_save_dir}/wm/{fs}/itr-{itr}.mp4'
+                save_results(pred_videos_1.cpu(),
+                             sample_video_file,
+                             fps=args.save_fps)
 
                 print('>' * 24)
-                # Decode segment and send to writer process
-                seg_video = model.decode_first_stage(
-                    wm_samples[:, :, :args.exe_steps]).detach().cpu()
-                write_q.put(seg_video)
+                # Collect the result of world-model interactions
+                wm_video.append(pred_videos_1[:, :, :args.exe_steps].cpu())
 
-            # Stop writer process
-            write_q.put(None)
-            writer_proc.join()
-
-    # Wait for all async I/O to complete
-    _flush_io()
+            full_video = torch.cat(wm_video, dim=2)
+            sample_tag = f"{args.dataset}-vid{sample['videoid']}-wd-fs-{fs}/full"
+            log_to_tensorboard(writer,
+                               full_video,
+                               sample_tag,
+                               fps=args.save_fps)
+            sample_full_video_file = f"{video_save_dir}/../{sample['videoid']}_full_fs{fs}.mp4"
+            save_results(full_video, sample_full_video_file, fps=args.save_fps)
 
 
 def get_parser():
@@ -946,11 +794,6 @@ def get_parser():
                         action='store_true',
                         default=False,
                         help="not using the predicted states as comparison")
-    parser.add_argument(
-        "--fast_policy_no_decode",
-        action='store_true',
-        default=True,
-        help="Speed mode: policy pass only predicts actions, skip policy video decode/log/save.")
     parser.add_argument("--save_fps",
                         type=int,
                         default=8,

scripts/trainer.py

@@ -11,9 +11,6 @@ from unifolm_wma.utils.utils import instantiate_from_config
 from unifolm_wma.utils.train import get_trainer_callbacks, get_trainer_logger, get_trainer_strategy
 from unifolm_wma.utils.train import set_logger, init_workspace, load_checkpoints, get_num_parameters
 
-torch.backends.cuda.matmul.allow_tf32 = True
-torch.backends.cudnn.allow_tf32 = True
-
 
 def get_parser(**parser_kwargs):
     parser = argparse.ArgumentParser(**parser_kwargs)

src/unifolm_wma/models/ddpms.py

@@ -988,7 +988,7 @@ class LatentDiffusion(DDPM):
 
     def instantiate_cond_stage(self, config: OmegaConf) -> None:
         """
-        Build the conditioning stage model. Frozen models are converted to FP16.
+        Build the conditioning stage model.
 
         Args:
             config: OmegaConf config describing the conditioning model to instantiate.
@@ -1000,7 +1000,6 @@ class LatentDiffusion(DDPM):
             self.cond_stage_model.train = disabled_train
             for param in self.cond_stage_model.parameters():
                 param.requires_grad = False
-            self.cond_stage_model.half()
         else:
             model = instantiate_from_config(config)
             self.cond_stage_model = model
@@ -1015,18 +1014,17 @@ class LatentDiffusion(DDPM):
         Returns:
             Conditioning embedding as a tensor (shape depends on cond model).
         """
-        with torch.cuda.amp.autocast(dtype=torch.float16):
-            if self.cond_stage_forward is None:
-                if hasattr(self.cond_stage_model, 'encode') and callable(
-                        self.cond_stage_model.encode):
-                    c = self.cond_stage_model.encode(c)
-                    if isinstance(c, DiagonalGaussianDistribution):
-                        c = c.mode()
-                else:
-                    c = self.cond_stage_model(c)
+        if self.cond_stage_forward is None:
+            if hasattr(self.cond_stage_model, 'encode') and callable(
+                    self.cond_stage_model.encode):
+                c = self.cond_stage_model.encode(c)
+                if isinstance(c, DiagonalGaussianDistribution):
+                    c = c.mode()
             else:
-                assert hasattr(self.cond_stage_model, self.cond_stage_forward)
-                c = getattr(self.cond_stage_model, self.cond_stage_forward)(c)
+                c = self.cond_stage_model(c)
+        else:
+            assert hasattr(self.cond_stage_model, self.cond_stage_forward)
+            c = getattr(self.cond_stage_model, self.cond_stage_forward)(c)
         return c
 
     def get_first_stage_encoding(
@@ -1959,7 +1957,6 @@ class LatentVisualDiffusion(LatentDiffusion):
             self.image_proj_model.train = disabled_train
             for param in self.image_proj_model.parameters():
                 param.requires_grad = False
-            self.image_proj_model.half()
 
     def _init_embedder(self, config: OmegaConf, freeze: bool = True) -> None:
         """
@@ -1975,7 +1972,6 @@ class LatentVisualDiffusion(LatentDiffusion):
             self.embedder.train = disabled_train
             for param in self.embedder.parameters():
                 param.requires_grad = False
-            self.embedder.half()
 
     def init_normalizers(self, normalize_config: OmegaConf,
                          dataset_stats: Mapping[str, Any]) -> None:
@@ -2179,9 +2175,8 @@ class LatentVisualDiffusion(LatentDiffusion):
             (random_num < 3 * self.uncond_prob).float(), "n -> n 1 1 1")
 
         cond_img = input_mask * img
-        with torch.cuda.amp.autocast(dtype=torch.float16):
-            cond_img_emb = self.embedder(cond_img)
-            cond_img_emb = self.image_proj_model(cond_img_emb)
+        cond_img_emb = self.embedder(cond_img)
+        cond_img_emb = self.image_proj_model(cond_img_emb)
 
         if self.model.conditioning_key == 'hybrid':
             if self.interp_mode:
@@ -2196,12 +2191,11 @@ class LatentVisualDiffusion(LatentDiffusion):
                                       repeat=z.shape[2])
             cond["c_concat"] = [img_cat_cond]
 
-        with torch.cuda.amp.autocast(dtype=torch.float16):
-            cond_action = self.action_projector(action)
-            cond_action_emb = self.agent_action_pos_emb + cond_action
-            # Get conditioning states
-            cond_state = self.state_projector(obs_state)
-            cond_state_emb = self.agent_state_pos_emb + cond_state
+        cond_action = self.action_projector(action)
+        cond_action_emb = self.agent_action_pos_emb + cond_action
+        # Get conditioning states
+        cond_state = self.state_projector(obs_state)
+        cond_state_emb = self.agent_state_pos_emb + cond_state
 
         if self.decision_making_only:
             is_sim_mode = False
@@ -2463,17 +2457,7 @@ class DiffusionWrapper(pl.LightningModule):
         Returns:
             Output from the inner diffusion model (tensor or tuple, depending on the model).
         """
-        with torch.cuda.amp.autocast(dtype=torch.float16):
-            return self._forward_impl(x, x_action, x_state, t,
-                                      c_concat, c_crossattn, c_crossattn_action,
-                                      c_adm, s, mask, **kwargs)
 
-    def _forward_impl(
-        self,
-        x, x_action, x_state, t,
-        c_concat=None, c_crossattn=None, c_crossattn_action=None,
-        c_adm=None, s=None, mask=None, **kwargs,
-    ):
         if self.conditioning_key is None:
             out = self.diffusion_model(x, t)
         elif self.conditioning_key == 'concat':

src/unifolm_wma/models/diffusion_head/conditional_unet1d.py

@@ -501,10 +501,6 @@ class ConditionalUnet1D(nn.Module):
         self.last_frame_only = last_frame_only
         self.horizon = horizon
 
-        # Context precomputation cache
-        self._global_cond_cache_enabled = False
-        self._global_cond_cache = {}
-
     def forward(self,
                 sample: torch.Tensor,
                 timestep: Union[torch.Tensor, float, int],
@@ -534,20 +530,14 @@ class ConditionalUnet1D(nn.Module):
         B, T, D = sample.shape
         if self.use_linear_act_proj:
             sample = self.proj_in_action(sample.unsqueeze(-1))
-            _gc_key = (cond['image'].data_ptr(), cond['agent_pos'].data_ptr())
-            if self._global_cond_cache_enabled and _gc_key in self._global_cond_cache:
-                global_cond = self._global_cond_cache[_gc_key]
-            else:
-                global_cond = self.obs_encoder(cond)
-                global_cond = rearrange(global_cond,
-                                        '(b t) d -> b 1 (t d)',
-                                        b=B,
-                                        t=self.n_obs_steps)
-                global_cond = repeat(global_cond,
-                                     'b c d -> b (repeat c) d',
-                                     repeat=T)
-                if self._global_cond_cache_enabled:
-                    self._global_cond_cache[_gc_key] = global_cond
+            global_cond = self.obs_encoder(cond)
+            global_cond = rearrange(global_cond,
+                                    '(b t) d -> b 1 (t d)',
+                                    b=B,
+                                    t=self.n_obs_steps)
+            global_cond = repeat(global_cond,
+                                 'b c d -> b (repeat c) d',
+                                 repeat=T)
         else:
             sample = einops.rearrange(sample, 'b h t -> b t h')
             sample = self.proj_in_horizon(sample)

src/unifolm_wma/models/samplers/ddim.py

@@ -6,8 +6,6 @@ from unifolm_wma.utils.diffusion import make_ddim_sampling_parameters, make_ddim
 from unifolm_wma.utils.common import noise_like
 from unifolm_wma.utils.common import extract_into_tensor
 from tqdm import tqdm
-from unifolm_wma.modules.attention import enable_cross_attn_kv_cache, disable_cross_attn_kv_cache
-from unifolm_wma.modules.networks.wma_model import enable_ctx_cache, disable_ctx_cache
 
 
 class DDIMSampler(object):
@@ -69,12 +67,11 @@ class DDIMSampler(object):
             ddim_timesteps=self.ddim_timesteps,
             eta=ddim_eta,
             verbose=verbose)
-        # Ensure tensors are on correct device for efficient indexing
-        self.register_buffer('ddim_sigmas', to_torch(torch.as_tensor(ddim_sigmas)))
-        self.register_buffer('ddim_alphas', to_torch(torch.as_tensor(ddim_alphas)))
-        self.register_buffer('ddim_alphas_prev', to_torch(torch.as_tensor(ddim_alphas_prev)))
+        self.register_buffer('ddim_sigmas', ddim_sigmas)
+        self.register_buffer('ddim_alphas', ddim_alphas)
+        self.register_buffer('ddim_alphas_prev', ddim_alphas_prev)
         self.register_buffer('ddim_sqrt_one_minus_alphas',
-                             to_torch(torch.as_tensor(np.sqrt(1. - ddim_alphas))))
+                             np.sqrt(1. - ddim_alphas))
         sigmas_for_original_sampling_steps = ddim_eta * torch.sqrt(
             (1 - self.alphas_cumprod_prev) / (1 - self.alphas_cumprod) *
             (1 - self.alphas_cumprod / self.alphas_cumprod_prev))
@@ -244,70 +241,63 @@ class DDIMSampler(object):
 
         dp_ddim_scheduler_action.set_timesteps(len(timesteps))
         dp_ddim_scheduler_state.set_timesteps(len(timesteps))
-        ts = torch.empty((b, ), device=device, dtype=torch.long)
-        enable_cross_attn_kv_cache(self.model)
-        enable_ctx_cache(self.model)
-        try:
-            for i, step in enumerate(iterator):
-                index = total_steps - i - 1
-                ts.fill_(step)
+        for i, step in enumerate(iterator):
+            index = total_steps - i - 1
+            ts = torch.full((b, ), step, device=device, dtype=torch.long)
 
-                # Use mask to blend noised original latent (img_orig) & new sampled latent (img)
-                if mask is not None:
-                    assert x0 is not None
-                    if clean_cond:
-                        img_orig = x0
-                    else:
-                        img_orig = self.model.q_sample(x0, ts)
-                    img = img_orig * mask + (1. - mask) * img
+            # Use mask to blend noised original latent (img_orig) & new sampled latent (img)
+            if mask is not None:
+                assert x0 is not None
+                if clean_cond:
+                    img_orig = x0
+                else:
+                    img_orig = self.model.q_sample(x0, ts)
+                img = img_orig * mask + (1. - mask) * img
 
-                outs = self.p_sample_ddim(
-                    img,
-                    action,
-                    state,
-                    cond,
-                    ts,
-                    index=index,
-                    use_original_steps=ddim_use_original_steps,
-                    quantize_denoised=quantize_denoised,
-                    temperature=temperature,
-                    noise_dropout=noise_dropout,
-                    score_corrector=score_corrector,
-                    corrector_kwargs=corrector_kwargs,
-                    unconditional_guidance_scale=unconditional_guidance_scale,
-                    unconditional_conditioning=unconditional_conditioning,
-                    mask=mask,
-                    x0=x0,
-                    fs=fs,
-                    guidance_rescale=guidance_rescale,
-                    **kwargs)
+            outs = self.p_sample_ddim(
+                img,
+                action,
+                state,
+                cond,
+                ts,
+                index=index,
+                use_original_steps=ddim_use_original_steps,
+                quantize_denoised=quantize_denoised,
+                temperature=temperature,
+                noise_dropout=noise_dropout,
+                score_corrector=score_corrector,
+                corrector_kwargs=corrector_kwargs,
+                unconditional_guidance_scale=unconditional_guidance_scale,
+                unconditional_conditioning=unconditional_conditioning,
+                mask=mask,
+                x0=x0,
+                fs=fs,
+                guidance_rescale=guidance_rescale,
+                **kwargs)
 
-                img, pred_x0, model_output_action, model_output_state = outs
+            img, pred_x0, model_output_action, model_output_state = outs
 
-                action = dp_ddim_scheduler_action.step(
-                    model_output_action,
-                    step,
-                    action,
-                    generator=None,
-                ).prev_sample
-                state = dp_ddim_scheduler_state.step(
-                    model_output_state,
-                    step,
-                    state,
-                    generator=None,
-                ).prev_sample
+            action = dp_ddim_scheduler_action.step(
+                model_output_action,
+                step,
+                action,
+                generator=None,
+            ).prev_sample
+            state = dp_ddim_scheduler_state.step(
+                model_output_state,
+                step,
+                state,
+                generator=None,
+            ).prev_sample
 
-                if callback: callback(i)
-                if img_callback: img_callback(pred_x0, i)
+            if callback: callback(i)
+            if img_callback: img_callback(pred_x0, i)
 
-                if index % log_every_t == 0 or index == total_steps - 1:
-                    intermediates['x_inter'].append(img)
-                    intermediates['pred_x0'].append(pred_x0)
-                    intermediates['x_inter_action'].append(action)
-                    intermediates['x_inter_state'].append(state)
-        finally:
-            disable_cross_attn_kv_cache(self.model)
-            disable_ctx_cache(self.model)
+            if index % log_every_t == 0 or index == total_steps - 1:
+                intermediates['x_inter'].append(img)
+                intermediates['pred_x0'].append(pred_x0)
+                intermediates['x_inter_action'].append(action)
+                intermediates['x_inter_state'].append(state)
 
         return img, action, state, intermediates
 
@@ -335,6 +325,10 @@ class DDIMSampler(object):
                       guidance_rescale=0.0,
                       **kwargs):
         b, *_, device = *x.shape, x.device
+        if x.dim() == 5:
+            is_video = True
+        else:
+            is_video = False
 
         if unconditional_conditioning is None or unconditional_guidance_scale == 1.:
             model_output, model_output_action, model_output_state = self.model.apply_model(
@@ -383,11 +377,17 @@ class DDIMSampler(object):
         sqrt_one_minus_alphas = self.model.sqrt_one_minus_alphas_cumprod if use_original_steps else self.ddim_sqrt_one_minus_alphas
         sigmas = self.ddim_sigmas_for_original_num_steps if use_original_steps else self.ddim_sigmas
 
-        # Use 0-d tensors directly (already on device); broadcasting handles shape
-        a_t = alphas[index]
-        a_prev = alphas_prev[index]
-        sigma_t = sigmas[index]
-        sqrt_one_minus_at = sqrt_one_minus_alphas[index]
+        if is_video:
+            size = (b, 1, 1, 1, 1)
+        else:
+            size = (b, 1, 1, 1)
+
+        a_t = torch.full(size, alphas[index], device=device)
+        a_prev = torch.full(size, alphas_prev[index], device=device)
+        sigma_t = torch.full(size, sigmas[index], device=device)
+        sqrt_one_minus_at = torch.full(size,
+                                       sqrt_one_minus_alphas[index],
+                                       device=device)
 
         if self.model.parameterization != "v":
             pred_x0 = (x - sqrt_one_minus_at * e_t) / a_t.sqrt()
@@ -395,8 +395,12 @@ class DDIMSampler(object):
             pred_x0 = self.model.predict_start_from_z_and_v(x, t, model_output)
 
         if self.model.use_dynamic_rescale:
-            scale_t = self.ddim_scale_arr[index]
-            prev_scale_t = self.ddim_scale_arr_prev[index]
+            scale_t = torch.full(size,
+                                 self.ddim_scale_arr[index],
+                                 device=device)
+            prev_scale_t = torch.full(size,
+                                      self.ddim_scale_arr_prev[index],
+                                      device=device)
             rescale = (prev_scale_t / scale_t)
             pred_x0 *= rescale
 

src/unifolm_wma/modules/attention.py

@@ -98,10 +98,6 @@ class CrossAttention(nn.Module):
         self.text_context_len = text_context_len
         self.agent_state_context_len = agent_state_context_len
         self.agent_action_context_len = agent_action_context_len
-        self._kv_cache = {}
-        self._kv_cache_enabled = False
-        self._kv_fused = False
-
         self.cross_attention_scale_learnable = cross_attention_scale_learnable
         if self.image_cross_attention:
             self.to_k_ip = nn.Linear(context_dim, inner_dim, bias=False)
@@ -118,27 +114,6 @@ class CrossAttention(nn.Module):
                 self.register_parameter('alpha_caa',
                                         nn.Parameter(torch.tensor(0.)))
 
-    def fuse_kv(self):
-        """Fuse to_k/to_v into to_kv (2 Linear → 1). Works for all layers."""
-        k_w = self.to_k.weight  # (inner_dim, context_dim)
-        v_w = self.to_v.weight
-        self.to_kv = nn.Linear(k_w.shape[1], k_w.shape[0] * 2, bias=False)
-        self.to_kv.weight = nn.Parameter(torch.cat([k_w, v_w], dim=0))
-        del self.to_k, self.to_v
-        if self.image_cross_attention:
-            for suffix in ('_ip', '_as', '_aa'):
-                k_attr = f'to_k{suffix}'
-                v_attr = f'to_v{suffix}'
-                kw = getattr(self, k_attr).weight
-                vw = getattr(self, v_attr).weight
-                fused = nn.Linear(kw.shape[1], kw.shape[0] * 2, bias=False)
-                fused.weight = nn.Parameter(torch.cat([kw, vw], dim=0))
-                setattr(self, f'to_kv{suffix}', fused)
-                delattr(self, k_attr)
-                delattr(self, v_attr)
-        self._kv_fused = True
-        return True
-
     def forward(self, x, context=None, mask=None):
         spatial_self_attn = (context is None)
         k_ip, v_ip, out_ip = None, None, None
@@ -165,28 +140,19 @@ class CrossAttention(nn.Module):
                                     self.agent_action_context_len +
                                     self.text_context_len:, :]
 
-            if self._kv_fused:
-                k, v = self.to_kv(context_ins).chunk(2, dim=-1)
-                k_ip, v_ip = self.to_kv_ip(context_image).chunk(2, dim=-1)
-                k_as, v_as = self.to_kv_as(context_agent_state).chunk(2, dim=-1)
-                k_aa, v_aa = self.to_kv_aa(context_agent_action).chunk(2, dim=-1)
-            else:
-                k = self.to_k(context_ins)
-                v = self.to_v(context_ins)
-                k_ip = self.to_k_ip(context_image)
-                v_ip = self.to_v_ip(context_image)
-                k_as = self.to_k_as(context_agent_state)
-                v_as = self.to_v_as(context_agent_state)
-                k_aa = self.to_k_aa(context_agent_action)
-                v_aa = self.to_v_aa(context_agent_action)
+            k = self.to_k(context_ins)
+            v = self.to_v(context_ins)
+            k_ip = self.to_k_ip(context_image)
+            v_ip = self.to_v_ip(context_image)
+            k_as = self.to_k_as(context_agent_state)
+            v_as = self.to_v_as(context_agent_state)
+            k_aa = self.to_k_aa(context_agent_action)
+            v_aa = self.to_v_aa(context_agent_action)
         else:
             if not spatial_self_attn:
                 context = context[:, :self.text_context_len, :]
-            if self._kv_fused:
-                k, v = self.to_kv(context).chunk(2, dim=-1)
-            else:
-                k = self.to_k(context)
-                v = self.to_v(context)
+            k = self.to_k(context)
+            v = self.to_v(context)
 
         q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h),
                       (q, k, v))
@@ -270,162 +236,134 @@ class CrossAttention(nn.Module):
         k_ip, v_ip, out_ip = None, None, None
         k_as, v_as, out_as = None, None, None
         k_aa, v_aa, out_aa = None, None, None
-        attn_mask_aa = None
 
-        h = self.heads
         q = self.to_q(x)
         context = default(context, x)
 
-        b, _, _ = q.shape
-        q = q.unsqueeze(3).reshape(b, q.shape[1], h, self.dim_head).permute(0, 2, 1, 3).reshape(b * h, q.shape[1], self.dim_head).contiguous()
-
-        def _reshape_kv(t):
-            return t.unsqueeze(3).reshape(b, t.shape[1], h, self.dim_head).permute(0, 2, 1, 3).reshape(b * h, t.shape[1], self.dim_head).contiguous()
-
-        use_cache = self._kv_cache_enabled and not spatial_self_attn
-        cache_hit = use_cache and len(self._kv_cache) > 0
-
-        if cache_hit:
-            k = self._kv_cache['k']
-            v = self._kv_cache['v']
-            k_ip = self._kv_cache.get('k_ip')
-            v_ip = self._kv_cache.get('v_ip')
-            k_as = self._kv_cache.get('k_as')
-            v_as = self._kv_cache.get('v_as')
-            k_aa = self._kv_cache.get('k_aa')
-            v_aa = self._kv_cache.get('v_aa')
-            attn_mask_aa = self._kv_cache.get('attn_mask_aa')
-        elif self.image_cross_attention and not spatial_self_attn:
+        if self.image_cross_attention and not spatial_self_attn:
             if context.shape[1] == self.text_context_len + self.video_length:
                 context_ins, context_image = context[:, :self.text_context_len, :], context[:,self.text_context_len:, :]
-                if self._kv_fused:
-                    k, v = self.to_kv(context).chunk(2, dim=-1)
-                    k_ip, v_ip = self.to_kv_ip(context_image).chunk(2, dim=-1)
-                else:
-                    k = self.to_k(context)
-                    v = self.to_v(context)
-                    k_ip = self.to_k_ip(context_image)
-                    v_ip = self.to_v_ip(context_image)
-                k, v = map(_reshape_kv, (k, v))
-                k_ip, v_ip = map(_reshape_kv, (k_ip, v_ip))
-                if use_cache:
-                    self._kv_cache = {'k': k, 'v': v, 'k_ip': k_ip, 'v_ip': v_ip}
+                k = self.to_k(context)
+                v = self.to_v(context)
+                k_ip = self.to_k_ip(context_image)
+                v_ip = self.to_v_ip(context_image)
             elif context.shape[1] == self.agent_state_context_len + self.text_context_len + self.video_length:
                 context_agent_state = context[:, :self.agent_state_context_len, :]
                 context_ins = context[:, self.agent_state_context_len:self.agent_state_context_len+self.text_context_len, :]
                 context_image = context[:, self.agent_state_context_len+self.text_context_len:, :]
-                if self._kv_fused:
-                    k, v = self.to_kv(context_ins).chunk(2, dim=-1)
-                    k_ip, v_ip = self.to_kv_ip(context_image).chunk(2, dim=-1)
-                    k_as, v_as = self.to_kv_as(context_agent_state).chunk(2, dim=-1)
-                else:
-                    k = self.to_k(context_ins)
-                    v = self.to_v(context_ins)
-                    k_ip = self.to_k_ip(context_image)
-                    v_ip = self.to_v_ip(context_image)
-                    k_as = self.to_k_as(context_agent_state)
-                    v_as = self.to_v_as(context_agent_state)
-                k, v = map(_reshape_kv, (k, v))
-                k_ip, v_ip = map(_reshape_kv, (k_ip, v_ip))
-                k_as, v_as = map(_reshape_kv, (k_as, v_as))
-                if use_cache:
-                    self._kv_cache = {'k': k, 'v': v, 'k_ip': k_ip, 'v_ip': v_ip, 'k_as': k_as, 'v_as': v_as}
+                k = self.to_k(context_ins)
+                v = self.to_v(context_ins)
+                k_ip = self.to_k_ip(context_image)
+                v_ip = self.to_v_ip(context_image)
+                k_as = self.to_k_as(context_agent_state)
+                v_as = self.to_v_as(context_agent_state)
             else:
                 context_agent_state = context[:, :self.agent_state_context_len, :]
                 context_agent_action = context[:, self.agent_state_context_len:self.agent_state_context_len+self.agent_action_context_len, :]
                 context_ins = context[:, self.agent_state_context_len+self.agent_action_context_len:self.agent_state_context_len+self.agent_action_context_len+self.text_context_len, :]
                 context_image = context[:, self.agent_state_context_len+self.agent_action_context_len+self.text_context_len:, :]
 
-                if self._kv_fused:
-                    k, v = self.to_kv(context_ins).chunk(2, dim=-1)
-                    k_ip, v_ip = self.to_kv_ip(context_image).chunk(2, dim=-1)
-                    k_as, v_as = self.to_kv_as(context_agent_state).chunk(2, dim=-1)
-                    k_aa, v_aa = self.to_kv_aa(context_agent_action).chunk(2, dim=-1)
-                else:
-                    k = self.to_k(context_ins)
-                    v = self.to_v(context_ins)
-                    k_ip = self.to_k_ip(context_image)
-                    v_ip = self.to_v_ip(context_image)
-                    k_as = self.to_k_as(context_agent_state)
-                    v_as = self.to_v_as(context_agent_state)
-                    k_aa = self.to_k_aa(context_agent_action)
-                    v_aa = self.to_v_aa(context_agent_action)
+                k = self.to_k(context_ins)
+                v = self.to_v(context_ins)
+                k_ip = self.to_k_ip(context_image)
+                v_ip = self.to_v_ip(context_image)
+                k_as = self.to_k_as(context_agent_state)
+                v_as = self.to_v_as(context_agent_state)
+                k_aa = self.to_k_aa(context_agent_action)
+                v_aa = self.to_v_aa(context_agent_action)
 
-                k, v = map(_reshape_kv, (k, v))
-                k_ip, v_ip = map(_reshape_kv, (k_ip, v_ip))
-                k_as, v_as = map(_reshape_kv, (k_as, v_as))
-                k_aa, v_aa = map(_reshape_kv, (k_aa, v_aa))
-
-                attn_mask_aa_raw = self._get_attn_mask_aa(x.shape[0],
-                                                          q.shape[1],
-                                                          k_aa.shape[1],
-                                                          block_size=16,
-                                                          device=k_aa.device)
-                attn_mask_aa = attn_mask_aa_raw.unsqueeze(1).repeat(1, h, 1, 1).reshape(
-                    b * h, attn_mask_aa_raw.shape[1], attn_mask_aa_raw.shape[2]).to(q.dtype)
-
-                if use_cache:
-                    self._kv_cache = {
-                        'k': k, 'v': v, 'k_ip': k_ip, 'v_ip': v_ip,
-                        'k_as': k_as, 'v_as': v_as, 'k_aa': k_aa, 'v_aa': v_aa,
-                        'attn_mask_aa': attn_mask_aa,
-                    }
+                attn_mask_aa = self._get_attn_mask_aa(x.shape[0],
+                                                      q.shape[1],
+                                                      k_aa.shape[1],
+                                                      block_size=16).to(k_aa.device)
         else:
             if not spatial_self_attn:
                 assert 1 > 2, ">>> ERROR: you should never go into here ..."
                 context = context[:, :self.text_context_len, :]
-            if self._kv_fused:
-                k, v = self.to_kv(context).chunk(2, dim=-1)
-            else:
-                k = self.to_k(context)
-                v = self.to_v(context)
-            k, v = map(_reshape_kv, (k, v))
-            if use_cache:
-                self._kv_cache = {'k': k, 'v': v}
+            k = self.to_k(context)
+            v = self.to_v(context)
+
+        b, _, _ = q.shape
+        q = q.unsqueeze(3).reshape(b, q.shape[1], self.heads, self.dim_head).permute(0, 2, 1, 3).reshape(b * self.heads, q.shape[1], self.dim_head).contiguous()
         if k is not None:
+            k, v = map(
+                lambda t: t.unsqueeze(3).reshape(b, t.shape[
+                    1], self.heads, self.dim_head).permute(0, 2, 1, 3).reshape(
+                        b * self.heads, t.shape[1], self.dim_head).contiguous(),
+                (k, v),
+            )
             out = xformers.ops.memory_efficient_attention(q,
                                                           k,
                                                           v,
                                                           attn_bias=None,
                                                           op=None)
             out = (out.unsqueeze(0).reshape(
-                b, h, out.shape[1],
+                b, self.heads, out.shape[1],
                 self.dim_head).permute(0, 2, 1,
                                        3).reshape(b, out.shape[1],
-                                                  h * self.dim_head))
+                                                  self.heads * self.dim_head))
 
         if k_ip is not None:
+            # For image cross-attention
+            k_ip, v_ip = map(
+                lambda t: t.unsqueeze(3).reshape(b, t.shape[
+                    1], self.heads, self.dim_head).permute(0, 2, 1, 3).reshape(
+                        b * self.heads, t.shape[1], self.dim_head).contiguous(
+                        ),
+                (k_ip, v_ip),
+            )
             out_ip = xformers.ops.memory_efficient_attention(q,
                                                              k_ip,
                                                              v_ip,
                                                              attn_bias=None,
                                                              op=None)
             out_ip = (out_ip.unsqueeze(0).reshape(
-                b, h, out_ip.shape[1],
+                b, self.heads, out_ip.shape[1],
                 self.dim_head).permute(0, 2, 1,
                                        3).reshape(b, out_ip.shape[1],
-                                                  h * self.dim_head))
+                                                  self.heads * self.dim_head))
 
         if k_as is not None:
+            # For agent state cross-attention
+            k_as, v_as = map(
+                lambda t: t.unsqueeze(3).reshape(b, t.shape[
+                    1], self.heads, self.dim_head).permute(0, 2, 1, 3).reshape(
+                        b * self.heads, t.shape[1], self.dim_head).contiguous(
+                        ),
+                (k_as, v_as),
+            )
             out_as = xformers.ops.memory_efficient_attention(q,
                                                              k_as,
                                                              v_as,
                                                              attn_bias=None,
                                                              op=None)
             out_as = (out_as.unsqueeze(0).reshape(
-                b, h, out_as.shape[1],
+                b, self.heads, out_as.shape[1],
                 self.dim_head).permute(0, 2, 1,
                                        3).reshape(b, out_as.shape[1],
-                                                  h * self.dim_head))
-
+                                                  self.heads * self.dim_head))
         if k_aa is not None:
+            # For agent action cross-attention
+            k_aa, v_aa = map(
+                lambda t: t.unsqueeze(3).reshape(b, t.shape[
+                    1], self.heads, self.dim_head).permute(0, 2, 1, 3).reshape(
+                        b * self.heads, t.shape[1], self.dim_head).contiguous(
+                        ),
+                (k_aa, v_aa),
+            )
+
+            attn_mask_aa = attn_mask_aa.unsqueeze(1).repeat(1,self.heads,1,1).reshape(
+                    b * self.heads, attn_mask_aa.shape[1], attn_mask_aa.shape[2])
+            attn_mask_aa = attn_mask_aa.to(q.dtype)
+
             out_aa = xformers.ops.memory_efficient_attention(
                 q, k_aa, v_aa, attn_bias=attn_mask_aa, op=None)
+
             out_aa = (out_aa.unsqueeze(0).reshape(
-                b, h, out_aa.shape[1],
+                b, self.heads, out_aa.shape[1],
                 self.dim_head).permute(0, 2, 1,
                                        3).reshape(b, out_aa.shape[1],
-                                                  h * self.dim_head))
+                                                  self.heads * self.dim_head))
         if exists(mask):
             raise NotImplementedError
 
@@ -448,43 +386,17 @@ class CrossAttention(nn.Module):
 
         return self.to_out(out)
 
-    def _get_attn_mask_aa(self, b, l1, l2, block_size=16, device=None):
-        cache_key = (b, l1, l2, block_size)
-        if hasattr(self, '_attn_mask_aa_cache_key') and self._attn_mask_aa_cache_key == cache_key:
-            cached = self._attn_mask_aa_cache
-            if device is not None and cached.device != torch.device(device):
-                cached = cached.to(device)
-                self._attn_mask_aa_cache = cached
-            return cached
-
-        target_device = device if device is not None else 'cpu'
+    def _get_attn_mask_aa(self, b, l1, l2, block_size=16):
         num_token = l2 // block_size
-        start_positions = ((torch.arange(b, device=target_device) % block_size) + 1) * num_token
-        col_indices = torch.arange(l2, device=target_device)
+        start_positions = ((torch.arange(b) % block_size) + 1) * num_token
+        col_indices = torch.arange(l2)
         mask_2d = col_indices.unsqueeze(0) >= start_positions.unsqueeze(1)
         mask = mask_2d.unsqueeze(1).expand(b, l1, l2)
-        attn_mask = torch.zeros(b, l1, l2, dtype=torch.float, device=target_device)
+        attn_mask = torch.zeros_like(mask, dtype=torch.float)
         attn_mask[mask] = float('-inf')
-
-        self._attn_mask_aa_cache_key = cache_key
-        self._attn_mask_aa_cache = attn_mask
         return attn_mask
 
 
-def enable_cross_attn_kv_cache(module):
-    for m in module.modules():
-        if isinstance(m, CrossAttention):
-            m._kv_cache_enabled = True
-            m._kv_cache = {}
-
-
-def disable_cross_attn_kv_cache(module):
-    for m in module.modules():
-        if isinstance(m, CrossAttention):
-            m._kv_cache_enabled = False
-            m._kv_cache = {}
-
-
 class BasicTransformerBlock(nn.Module):
 
     def __init__(self,

src/unifolm_wma/modules/networks/wma_model.py

@@ -685,21 +685,6 @@ class WMAModel(nn.Module):
         self.action_token_projector = instantiate_from_config(
             stem_process_config)
 
-        # Context precomputation cache
-        self._ctx_cache_enabled = False
-        self._ctx_cache = {}
-        # Reusable CUDA stream for parallel state_unet / action_unet
-        self._state_stream = torch.cuda.Stream()
-
-    def __getstate__(self):
-        state = self.__dict__.copy()
-        state.pop('_state_stream', None)
-        return state
-
-    def __setstate__(self, state):
-        self.__dict__.update(state)
-        self._state_stream = torch.cuda.Stream()
-
     def forward(self,
                 x: Tensor,
                 x_action: Tensor,
@@ -735,64 +720,58 @@ class WMAModel(nn.Module):
                                    repeat_only=False).type(x.dtype)
         emb = self.time_embed(t_emb)
 
-        _ctx_key = context.data_ptr()
-        if self._ctx_cache_enabled and _ctx_key in self._ctx_cache:
-            context = self._ctx_cache[_ctx_key]
+        bt, l_context, _ = context.shape
+        if self.base_model_gen_only:
+            assert l_context == 77 + self.n_obs_steps * 16, ">>> ERROR Context dim 1 ..."  ## NOTE HANDCODE
         else:
-            bt, l_context, _ = context.shape
-            if self.base_model_gen_only:
-                assert l_context == 77 + self.n_obs_steps * 16, ">>> ERROR Context dim 1 ..."  ## NOTE HANDCODE
-            else:
-                if l_context == self.n_obs_steps + 77 + t * 16:
-                    context_agent_state = context[:, :self.n_obs_steps]
-                    context_text = context[:, self.n_obs_steps:self.n_obs_steps +
-                                           77, :]
-                    context_img = context[:, self.n_obs_steps + 77:, :]
-                    context_agent_state = context_agent_state.repeat_interleave(
-                        repeats=t, dim=0)
-                    context_text = context_text.repeat_interleave(repeats=t, dim=0)
-                    context_img = rearrange(context_img,
-                                            'b (t l) c -> (b t) l c',
-                                            t=t)
-                    context = torch.cat(
-                        [context_agent_state, context_text, context_img], dim=1)
-                elif l_context == self.n_obs_steps + 16 + 77 + t * 16:
-                    context_agent_state = context[:, :self.n_obs_steps]
-                    context_agent_action = context[:, self.
-                                                   n_obs_steps:self.n_obs_steps +
-                                                   16, :]
-                    context_agent_action = rearrange(
-                        context_agent_action.unsqueeze(2), 'b t l d -> (b t) l d')
-                    context_agent_action = self.action_token_projector(
-                        context_agent_action)
-                    context_agent_action = rearrange(context_agent_action,
-                                                     '(b o) l d -> b o l d',
-                                                     o=t)
-                    context_agent_action = rearrange(context_agent_action,
-                                                     'b o (t l) d -> b o t l d',
-                                                     t=t)
-                    context_agent_action = context_agent_action.permute(
-                        0, 2, 1, 3, 4)
-                    context_agent_action = rearrange(context_agent_action,
-                                                     'b t o l d -> (b t) (o l) d')
+            if l_context == self.n_obs_steps + 77 + t * 16:
+                context_agent_state = context[:, :self.n_obs_steps]
+                context_text = context[:, self.n_obs_steps:self.n_obs_steps +
+                                       77, :]
+                context_img = context[:, self.n_obs_steps + 77:, :]
+                context_agent_state = context_agent_state.repeat_interleave(
+                    repeats=t, dim=0)
+                context_text = context_text.repeat_interleave(repeats=t, dim=0)
+                context_img = rearrange(context_img,
+                                        'b (t l) c -> (b t) l c',
+                                        t=t)
+                context = torch.cat(
+                    [context_agent_state, context_text, context_img], dim=1)
+            elif l_context == self.n_obs_steps + 16 + 77 + t * 16:
+                context_agent_state = context[:, :self.n_obs_steps]
+                context_agent_action = context[:, self.
+                                               n_obs_steps:self.n_obs_steps +
+                                               16, :]
+                context_agent_action = rearrange(
+                    context_agent_action.unsqueeze(2), 'b t l d -> (b t) l d')
+                context_agent_action = self.action_token_projector(
+                    context_agent_action)
+                context_agent_action = rearrange(context_agent_action,
+                                                 '(b o) l d -> b o l d',
+                                                 o=t)
+                context_agent_action = rearrange(context_agent_action,
+                                                 'b o (t l) d -> b o t l d',
+                                                 t=t)
+                context_agent_action = context_agent_action.permute(
+                    0, 2, 1, 3, 4)
+                context_agent_action = rearrange(context_agent_action,
+                                                 'b t o l d -> (b t) (o l) d')
 
-                    context_text = context[:, self.n_obs_steps +
-                                           16:self.n_obs_steps + 16 + 77, :]
-                    context_text = context_text.repeat_interleave(repeats=t, dim=0)
+                context_text = context[:, self.n_obs_steps +
+                                       16:self.n_obs_steps + 16 + 77, :]
+                context_text = context_text.repeat_interleave(repeats=t, dim=0)
 
-                    context_img = context[:, self.n_obs_steps + 16 + 77:, :]
-                    context_img = rearrange(context_img,
-                                            'b (t l) c -> (b t) l c',
-                                            t=t)
-                    context_agent_state = context_agent_state.repeat_interleave(
-                        repeats=t, dim=0)
-                    context = torch.cat([
-                        context_agent_state, context_agent_action, context_text,
-                        context_img
-                    ],
-                                        dim=1)
-            if self._ctx_cache_enabled:
-                self._ctx_cache[_ctx_key] = context
+                context_img = context[:, self.n_obs_steps + 16 + 77:, :]
+                context_img = rearrange(context_img,
+                                        'b (t l) c -> (b t) l c',
+                                        t=t)
+                context_agent_state = context_agent_state.repeat_interleave(
+                    repeats=t, dim=0)
+                context = torch.cat([
+                    context_agent_state, context_agent_action, context_text,
+                    context_img
+                ],
+                                    dim=1)
 
         emb = emb.repeat_interleave(repeats=t, dim=0)
 
@@ -853,45 +832,17 @@ class WMAModel(nn.Module):
 
         if not self.base_model_gen_only:
             ba, _, _ = x_action.shape
-            ts_state = timesteps[:ba] if b > 1 else timesteps
-            # Run action_unet and state_unet in parallel via CUDA streams
-            s_stream = self._state_stream
-            s_stream.wait_stream(torch.cuda.current_stream())
-            with torch.cuda.stream(s_stream):
-                s_y = self.state_unet(x_state, ts_state, hs_a,
-                                      context_action[:2], **kwargs)
             a_y = self.action_unet(x_action, timesteps[:ba], hs_a,
                                    context_action[:2], **kwargs)
-            torch.cuda.current_stream().wait_stream(s_stream)
+            # Predict state
+            if b > 1:
+                s_y = self.state_unet(x_state, timesteps[:ba], hs_a,
+                                      context_action[:2], **kwargs)
+            else:
+                s_y = self.state_unet(x_state, timesteps, hs_a,
+                                      context_action[:2], **kwargs)
         else:
             a_y = torch.zeros_like(x_action)
             s_y = torch.zeros_like(x_state)
 
         return y, a_y, s_y
-
-
-def enable_ctx_cache(model):
-    """Enable context precomputation cache on WMAModel and its action/state UNets."""
-    for m in model.modules():
-        if isinstance(m, WMAModel):
-            m._ctx_cache_enabled = True
-            m._ctx_cache = {}
-    # conditional_unet1d cache
-    from unifolm_wma.models.diffusion_head.conditional_unet1d import ConditionalUnet1D
-    for m in model.modules():
-        if isinstance(m, ConditionalUnet1D):
-            m._global_cond_cache_enabled = True
-            m._global_cond_cache = {}
-
-
-def disable_ctx_cache(model):
-    """Disable and clear context precomputation cache."""
-    for m in model.modules():
-        if isinstance(m, WMAModel):
-            m._ctx_cache_enabled = False
-            m._ctx_cache = {}
-    from unifolm_wma.models.diffusion_head.conditional_unet1d import ConditionalUnet1D
-    for m in model.modules():
-        if isinstance(m, ConditionalUnet1D):
-            m._global_cond_cache_enabled = False
-            m._global_cond_cache = {}

unitree_g1_pack_camera/case1/output.log

@@ -1,16 +1,21 @@
-2026-02-19 18:55:32.160020: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 18:55:32.207538: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 18:55:32.207581: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 18:55:32.208613: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 18:55:32.215249: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 19:49:03.885238: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 19:49:03.934263: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 19:49:03.934309: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 19:49:03.935622: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 19:49:03.943041: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 18:55:33.121466: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 19:49:04.852993: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:12<12:09, 72.95s/it]
  18%|█▊        | 2/11 [02:26<10:58, 73.19s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_g1_pack_camera/case1/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_g1_pack_camera/case1/unitree_g1_pack_camera_case1.mp4",
     "pred_video": "unitree_g1_pack_camera/case1/output/inference/0_full_fs6.mp4",
-    "psnr": 32.34126103448495
+    "psnr": 35.615362167470806
 }

unitree_g1_pack_camera/case1/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_g1_pack_camera"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_g1_pack_camera/case2/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:00:05.944067: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:00:05.991354: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:00:05.991392: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:00:05.992414: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:00:05.999050: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 20:04:06.049535: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 20:04:06.099186: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 20:04:06.099232: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 20:04:06.100544: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 20:04:06.108023: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:00:06.916175: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 20:04:07.025500: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:14<12:22, 74.22s/it]
  18%|█▊        | 2/11 [02:28<11:09, 74.33s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_g1_pack_camera/case2/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_g1_pack_camera/case2/unitree_g1_pack_camera_case2.mp4",
     "pred_video": "unitree_g1_pack_camera/case2/output/inference/50_full_fs6.mp4",
-    "psnr": 37.49178506869336
+    "psnr": 34.61979248212279
 }

unitree_g1_pack_camera/case2/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_g1_pack_camera"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_g1_pack_camera/case3/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:04:41.036634: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:04:41.084414: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:04:41.084452: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:04:41.085481: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:04:41.092287: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 20:19:19.271045: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 20:19:19.320688: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 20:19:19.320734: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 20:19:19.322059: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 20:19:19.329606: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:04:42.000614: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 20:19:20.248938: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:14<12:22, 74.28s/it]
  18%|█▊        | 2/11 [02:28<11:09, 74.38s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_g1_pack_camera/case3/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_g1_pack_camera/case3/unitree_g1_pack_camera_case3.mp4",
     "pred_video": "unitree_g1_pack_camera/case3/output/inference/100_full_fs6.mp4",
-    "psnr": 29.88155122131729
+    "psnr": 37.034952654534486
 }

unitree_g1_pack_camera/case3/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_g1_pack_camera"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_g1_pack_camera/case4/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:09:16.122268: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:09:16.170290: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:09:16.170331: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:09:16.171349: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:09:16.177993: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 20:34:34.563818: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 20:34:34.613426: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 20:34:34.613485: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 20:34:34.614802: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 20:34:34.622286: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:09:17.087425: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 20:34:35.540506: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:14<12:25, 74.52s/it]
  18%|█▊        | 2/11 [02:29<11:15, 75.00s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_g1_pack_camera/case4/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_g1_pack_camera/case4/unitree_g1_pack_camera_case4.mp4",
     "pred_video": "unitree_g1_pack_camera/case4/output/inference/200_full_fs6.mp4",
-    "psnr": 35.62512454155058
+    "psnr": 31.43390896360405
 }

unitree_g1_pack_camera/case4/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_g1_pack_camera"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_cleanup_pencils/case1/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:13:51.554194: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:13:51.601580: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:13:51.601622: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:13:51.602646: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:13:51.609297: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 20:49:47.965949: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 20:49:48.015942: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 20:49:48.015997: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 20:49:48.017330: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 20:49:48.024854: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:13:52.517676: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 20:49:48.943205: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/8 [00:00<?, ?it/s]
+
   0%|          | 0/8 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  12%|█▎        | 1/8 [01:15<08:48, 75.51s/it]
  25%|██▌       | 2/8 [02:30<07:32, 75.39s/it]
@@ -60,6 +113,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 4: generating actions ...
 >>> Step 4: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
->>> Step 5: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...
+>>> Step 5: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_cleanup_pencils/case1/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_cleanup_pencils/case1/unitree_z1_dual_arm_cleanup_pencils_case1.mp4",
     "pred_video": "unitree_z1_dual_arm_cleanup_pencils/case1/output/inference/0_full_fs4.mp4",
-    "psnr": 38.269577028444445
+    "psnr": 47.911564449209735
 }

unitree_z1_dual_arm_cleanup_pencils/case1/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_cleanup_pencils"
         --n_iter 8 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_cleanup_pencils/case2/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:17:16.282875: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:17:16.330519: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:17:16.330561: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:17:16.331631: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:17:16.338413: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 21:01:19.535243: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 21:01:19.585230: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 21:01:19.585275: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 21:01:19.586600: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 21:01:19.594107: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:17:17.250653: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 21:01:20.510688: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/8 [00:00<?, ?it/s]
+
   0%|          | 0/8 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  12%|█▎        | 1/8 [01:16<08:54, 76.34s/it]
  25%|██▌       | 2/8 [02:32<07:37, 76.28s/it]
@@ -60,6 +113,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 4: generating actions ...
 >>> Step 4: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
->>> Step 5: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...
+>>> Step 5: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_cleanup_pencils/case2/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_cleanup_pencils/case2/unitree_z1_dual_arm_cleanup_pencils_case2.mp4",
     "pred_video": "unitree_z1_dual_arm_cleanup_pencils/case2/output/inference/50_full_fs4.mp4",
-    "psnr": 44.50028075962896
+    "psnr": 48.344571927558974
 }

unitree_z1_dual_arm_cleanup_pencils/case2/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_cleanup_pencils"
         --n_iter 8 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_cleanup_pencils/case3/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:20:40.444703: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:20:40.492237: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:20:40.492278: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:20:40.493360: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:20:40.500130: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 21:13:04.812376: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 21:13:04.862167: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 21:13:04.862223: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 21:13:04.863549: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 21:13:04.871078: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:20:41.414718: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 21:13:05.785070: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/8 [00:00<?, ?it/s]
+
   0%|          | 0/8 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  12%|█▎        | 1/8 [01:15<08:45, 75.11s/it]
  25%|██▌       | 2/8 [02:30<07:31, 75.30s/it]
@@ -60,6 +113,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 4: generating actions ...
 >>> Step 4: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
->>> Step 5: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...
+>>> Step 5: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_cleanup_pencils/case3/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_cleanup_pencils/case3/unitree_z1_dual_arm_cleanup_pencils_case3.mp4",
     "pred_video": "unitree_z1_dual_arm_cleanup_pencils/case3/output/inference/100_full_fs4.mp4",
-    "psnr": 32.29959078097713
+    "psnr": 41.152374490134825
 }

unitree_z1_dual_arm_cleanup_pencils/case3/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_cleanup_pencils"
         --n_iter 8 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_cleanup_pencils/case4/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:24:05.230366: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:24:05.278058: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:24:05.278100: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:24:05.279133: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:24:05.285789: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 21:24:42.443699: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 21:24:42.494143: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 21:24:42.494201: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 21:24:42.495506: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 21:24:42.503003: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:24:06.199101: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 21:24:43.415898: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/8 [00:00<?, ?it/s]
+
   0%|          | 0/8 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  12%|█▎        | 1/8 [01:15<08:46, 75.28s/it]
  25%|██▌       | 2/8 [02:30<07:32, 75.34s/it]
@@ -60,6 +113,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 4: generating actions ...
 >>> Step 4: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
->>> Step 5: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...
+>>> Step 5: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_cleanup_pencils/case4/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_cleanup_pencils/case4/unitree_z1_dual_arm_cleanup_pencils_case4.mp4",
     "pred_video": "unitree_z1_dual_arm_cleanup_pencils/case4/output/inference/200_full_fs4.mp4",
-    "psnr": 45.051241961122535
+    "psnr": 46.025723557253855
 }

unitree_z1_dual_arm_cleanup_pencils/case4/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_cleanup_pencils"
         --n_iter 8 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox/case1/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:27:29.317502: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:27:29.365030: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:27:29.365079: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:27:29.366111: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:27:29.372733: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 21:36:14.761055: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 21:36:14.811056: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 21:36:14.811115: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 21:36:14.812480: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 21:36:14.820115: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:27:30.291220: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 21:36:15.736583: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/7 [00:00<?, ?it/s]
+
   0%|          | 0/7 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  14%|█▍        | 1/7 [01:15<07:34, 75.70s/it]
  29%|██▊       | 2/7 [02:31<06:18, 75.65s/it]
@@ -57,6 +110,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 3: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 4: generating actions ...
->>> Step 4: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
+>>> Step 4: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...

unitree_z1_dual_arm_stackbox/case1/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox/case1/unitree_z1_dual_arm_stackbox_case1.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox/case1/output/inference/5_full_fs4.mp4",
-    "psnr": 42.717688631296596
+    "psnr": 44.3480149502738
 }

unitree_z1_dual_arm_stackbox/case1/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox"
         --n_iter 7 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox/case2/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:30:30.058862: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:30:30.106200: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:30:30.106243: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:30:30.107276: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:30:30.113917: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 21:46:41.375935: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 21:46:41.426557: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 21:46:41.426614: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 21:46:41.427937: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 21:46:41.435507: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:30:31.026240: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 21:46:42.361310: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/7 [00:00<?, ?it/s]
+
   0%|          | 0/7 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  14%|█▍        | 1/7 [01:16<07:38, 76.39s/it]
  29%|██▊       | 2/7 [02:33<06:23, 76.69s/it]
@@ -57,6 +110,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 3: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 4: generating actions ...
->>> Step 4: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
+>>> Step 4: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...

unitree_z1_dual_arm_stackbox/case2/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox/case2/unitree_z1_dual_arm_stackbox_case2.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox/case2/output/inference/15_full_fs4.mp4",
-    "psnr": 44.90750363879194
+    "psnr": 39.867728254007716
 }

unitree_z1_dual_arm_stackbox/case2/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox"
         --n_iter 7 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox/case3/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:33:31.235859: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:33:31.283866: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:33:31.283908: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:33:31.284941: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:33:31.291610: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 21:57:17.623993: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 21:57:17.673835: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 21:57:17.673891: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 21:57:17.675211: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 21:57:17.682716: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:33:32.199716: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 21:57:18.593525: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/7 [00:00<?, ?it/s]
+
   0%|          | 0/7 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  14%|█▍        | 1/7 [01:15<07:33, 75.59s/it]
  29%|██▊       | 2/7 [02:31<06:17, 75.59s/it]
@@ -57,6 +110,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 3: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 4: generating actions ...
->>> Step 4: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
+>>> Step 4: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...

unitree_z1_dual_arm_stackbox/case3/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox/case3/unitree_z1_dual_arm_stackbox_case3.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox/case3/output/inference/25_full_fs4.mp4",
-    "psnr": 39.63695040491171
+    "psnr": 39.19101039445159
 }

unitree_z1_dual_arm_stackbox/case3/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox"
         --n_iter 7 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox/case4/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:36:32.251051: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:36:32.298464: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:36:32.298506: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:36:32.299538: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:36:32.306168: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 22:07:43.398736: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 22:07:43.448264: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 22:07:43.448321: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 22:07:43.449636: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 22:07:43.457127: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:36:33.213503: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 22:07:44.370935: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/7 [00:00<?, ?it/s]
+
   0%|          | 0/7 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
  14%|█▍        | 1/7 [01:17<07:42, 77.04s/it]
  29%|██▊       | 2/7 [02:33<06:24, 76.95s/it]
@@ -57,6 +110,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 3: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 4: generating actions ...
->>> Step 4: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 5: generating actions ...
+>>> Step 4: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 5: generating actions ...

unitree_z1_dual_arm_stackbox/case4/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox/case4/unitree_z1_dual_arm_stackbox_case4.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox/case4/output/inference/35_full_fs4.mp4",
-    "psnr": 42.34177660061245
+    "psnr": 40.29563315341769
 }

unitree_z1_dual_arm_stackbox/case4/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox"
         --n_iter 7 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox_v2/case1/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:39:32.908698: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:39:32.956378: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:39:32.956417: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:39:32.957459: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:39:32.964104: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 22:18:17.396072: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 22:18:17.446095: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 22:18:17.446154: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 22:18:17.447480: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 22:18:17.455025: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:39:33.875854: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 22:18:18.367007: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:15<12:35, 75.53s/it]
  18%|█▊        | 2/11 [02:30<11:18, 75.39s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_stackbox_v2/case1/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox_v2/case1/unitree_z1_dual_arm_stackbox_v2_case1.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox_v2/case1/output/inference/5_full_fs4.mp4",
-    "psnr": 26.68301835085306
+    "psnr": 25.812741419225095
 }

unitree_z1_dual_arm_stackbox_v2/case1/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox_v2"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox_v2/case2/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:44:07.724109: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:44:07.771461: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:44:07.771505: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:44:07.772537: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:44:07.779172: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 22:33:42.261398: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 22:33:42.310786: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 22:33:42.310845: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 22:33:42.312191: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 22:33:42.319738: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:44:08.688975: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 22:33:43.232517: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:13<12:18, 73.90s/it]
  18%|█▊        | 2/11 [02:27<11:05, 73.99s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_stackbox_v2/case2/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox_v2/case2/unitree_z1_dual_arm_stackbox_v2_case2.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox_v2/case2/output/inference/15_full_fs4.mp4",
-    "psnr": 27.46347145461597
+    "psnr": 33.90444714332389
 }

unitree_z1_dual_arm_stackbox_v2/case2/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox_v2"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox_v2/case3/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:48:42.460586: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:48:42.508096: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:48:42.508140: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:48:42.509152: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:48:42.515865: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 22:48:49.761688: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 22:48:49.811395: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 22:48:49.811456: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 22:48:49.812798: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 22:48:49.820307: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:48:43.425699: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 22:48:50.732941: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:14<12:29, 74.99s/it]
  18%|█▊        | 2/11 [02:30<11:18, 75.44s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_stackbox_v2/case3/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox_v2/case3/unitree_z1_dual_arm_stackbox_v2_case3.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox_v2/case3/output/inference/25_full_fs4.mp4",
-    "psnr": 28.604047286947512
+    "psnr": 34.50192428908007
 }

unitree_z1_dual_arm_stackbox_v2/case3/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox_v2"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_dual_arm_stackbox_v2/case4/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:53:17.574354: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:53:17.621335: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:53:17.621388: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:53:17.622415: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:53:17.629050: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 23:04:15.762959: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 23:04:15.814243: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 23:04:15.814301: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 23:04:15.815653: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 23:04:15.823287: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:53:18.537233: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 23:04:16.742609: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/11 [00:00<?, ?it/s]
+
   0%|          | 0/11 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   9%|▉         | 1/11 [01:14<12:25, 74.55s/it]
  18%|█▊        | 2/11 [02:29<11:13, 74.83s/it]
@@ -69,6 +122,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>> Step 6: generating actions ...
 >>> Step 6: interacting with world model ...
 >>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 7: generating actions ...
->>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 7: generating actions ...
+>>> Step 7: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>

unitree_z1_dual_arm_stackbox_v2/case4/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_dual_arm_stackbox_v2/case4/unitree_z1_dual_arm_stackbox_v2_case4.mp4",
     "pred_video": "unitree_z1_dual_arm_stackbox_v2/case4/output/inference/35_full_fs4.mp4",
-    "psnr": 25.578757174083307
+    "psnr": 38.797893493652516
 }

unitree_z1_dual_arm_stackbox_v2/case4/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_dual_arm_stackbox_v2"
         --n_iter 11 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_stackbox/case1/output.log

@@ -1,16 +1,21 @@
-2026-02-19 19:57:52.488339: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 19:57:52.536176: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 19:57:52.536222: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 19:57:52.537285: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 19:57:52.544051: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 23:19:36.475817: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 23:19:36.525118: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 23:19:36.525172: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 23:19:36.526479: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 23:19:36.533981: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 19:57:53.469912: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 23:19:37.461985: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/12 [00:00<?, ?it/s]
+
   0%|          | 0/12 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   8%|▊         | 1/12 [01:14<13:38, 74.37s/it]
  17%|█▋        | 2/12 [02:28<12:21, 74.13s/it]
@@ -72,6 +125,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 7: generating actions ...
 >>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 8: generating actions ...
->>> Step 8: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 8: generating actions ...
+>>> Step 8: interacting with world model ...

unitree_z1_stackbox/case1/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_stackbox/case1/unitree_z1_stackbox_case1.mp4",
     "pred_video": "unitree_z1_stackbox/case1/output/inference/5_full_fs4.mp4",
-    "psnr": 46.05271283048069
+    "psnr": 42.83913947323794
 }

unitree_z1_stackbox/case1/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_stackbox"
         --n_iter 12 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_stackbox/case2/output.log

@@ -1,16 +1,21 @@
-2026-02-19 20:02:50.975402: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 20:02:51.023211: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 20:02:51.023253: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 20:02:51.024328: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 20:02:51.031176: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 23:35:52.961572: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 23:35:53.011562: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 23:35:53.011622: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 23:35:53.012961: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 23:35:53.020520: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 20:02:51.947400: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 23:35:53.953177: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/12 [00:00<?, ?it/s]
+
   0%|          | 0/12 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   8%|▊         | 1/12 [01:13<13:33, 73.92s/it]
  17%|█▋        | 2/12 [02:28<12:22, 74.25s/it]
@@ -72,6 +125,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 7: generating actions ...
 >>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 8: generating actions ...
->>> Step 8: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 8: generating actions ...
+>>> Step 8: interacting with world model ...

unitree_z1_stackbox/case2/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_stackbox/case2/unitree_z1_stackbox_case2.mp4",
     "pred_video": "unitree_z1_stackbox/case2/output/inference/15_full_fs4.mp4",
-    "psnr": 43.005233352958804
+    "psnr": 48.64571989587276
 }

unitree_z1_stackbox/case2/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_stackbox"
         --n_iter 12 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_stackbox/case3/output.log

@@ -1,16 +1,21 @@
-2026-02-19 20:07:49.410622: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 20:07:49.457896: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 20:07:49.457948: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 20:07:49.458967: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 20:07:49.465632: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-11 23:51:59.440508: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-11 23:51:59.489849: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-11 23:51:59.489897: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-11 23:51:59.491194: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-11 23:51:59.498659: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 20:07:50.373326: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-11 23:52:00.422248: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/12 [00:00<?, ?it/s]
+
   0%|          | 0/12 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   8%|▊         | 1/12 [01:14<13:34, 74.06s/it]
  17%|█▋        | 2/12 [02:27<12:17, 73.74s/it]
@@ -72,6 +125,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 7: generating actions ...
 >>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 8: generating actions ...
->>> Step 8: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 8: generating actions ...
+>>> Step 8: interacting with world model ...

unitree_z1_stackbox/case3/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_stackbox/case3/unitree_z1_stackbox_case3.mp4",
     "pred_video": "unitree_z1_stackbox/case3/output/inference/25_full_fs4.mp4",
-    "psnr": 49.489774674892764
+    "psnr": 45.127553229898034
 }

unitree_z1_stackbox/case3/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_stackbox"
         --n_iter 12 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"

unitree_z1_stackbox/case4/output.log

@@ -1,16 +1,21 @@
-2026-02-19 20:12:48.029611: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
-2026-02-19 20:12:48.076914: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
-2026-02-19 20:12:48.076957: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
-2026-02-19 20:12:48.077981: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
-2026-02-19 20:12:48.084620: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
+2026-02-12 00:08:05.532356: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
+2026-02-12 00:08:05.582134: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
+2026-02-12 00:08:05.582180: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
+2026-02-12 00:08:05.583503: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
+2026-02-12 00:08:05.591049: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
 To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
-2026-02-19 20:12:49.004753: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
+2026-02-12 00:08:06.523218: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
 Global seed set to 123
->>> Loading prepared model from ckpts/unifolm_wma_dual.ckpt.prepared.pt ...
->>> Prepared model loaded.
->>> Diffusion backbone (model.model) converted to FP16.
->>> Projectors (image_proj_model, state_projector, action_projector) converted to FP16.
->>> Encoders (cond_stage_model, embedder) converted to FP16.
+INFO:mainlogger:LatentVisualDiffusion: Running in v-prediction mode
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+INFO:unifolm_wma.models.diffusion_head.conditional_unet1d:number of parameters: 5.010531e+08
+AE working on z of shape (1, 4, 32, 32) = 4096 dimensions.
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+INFO:root:Loaded ViT-H-14 model config.
+INFO:root:Loading pretrained ViT-H-14 weights (laion2b_s32b_b79k).
+>>> model checkpoint loaded.
+>>> Load pre-trained model ...
 INFO:root:***** Configing Data *****
 >>> unitree_z1_stackbox: 1 data samples loaded.
 >>> unitree_z1_stackbox: data stats loaded.
@@ -28,15 +33,63 @@ INFO:root:***** Configing Data *****
 >>> unitree_g1_pack_camera: data stats loaded.
 >>> unitree_g1_pack_camera: normalizer initiated.
 >>> Dataset is successfully loaded ...
-    ✓ KV fused: 66 attention layers
 >>> Generate 16 frames under each generation ...
 DEBUG:h5py._conv:Creating converter from 3 to 5
 DEBUG:PIL.PngImagePlugin:STREAM b'IHDR' 16 13
 DEBUG:PIL.PngImagePlugin:STREAM b'pHYs' 41 9
 DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
-
-  0%|          | 0/12 [00:00<?, ?it/s]
+
   0%|          | 0/12 [00:00<?, ?it/s]>>> Step 0: generating actions ...
+>>> Step 0: interacting with world model ...
+DEBUG:PIL.Image:Importing BlpImagePlugin
+DEBUG:PIL.Image:Importing BmpImagePlugin
+DEBUG:PIL.Image:Importing BufrStubImagePlugin
+DEBUG:PIL.Image:Importing CurImagePlugin
+DEBUG:PIL.Image:Importing DcxImagePlugin
+DEBUG:PIL.Image:Importing DdsImagePlugin
+DEBUG:PIL.Image:Importing EpsImagePlugin
+DEBUG:PIL.Image:Importing FitsImagePlugin
+DEBUG:PIL.Image:Importing FitsStubImagePlugin
+DEBUG:PIL.Image:Importing FliImagePlugin
+DEBUG:PIL.Image:Importing FpxImagePlugin
+DEBUG:PIL.Image:Image: failed to import FpxImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing FtexImagePlugin
+DEBUG:PIL.Image:Importing GbrImagePlugin
+DEBUG:PIL.Image:Importing GifImagePlugin
+DEBUG:PIL.Image:Importing GribStubImagePlugin
+DEBUG:PIL.Image:Importing Hdf5StubImagePlugin
+DEBUG:PIL.Image:Importing IcnsImagePlugin
+DEBUG:PIL.Image:Importing IcoImagePlugin
+DEBUG:PIL.Image:Importing ImImagePlugin
+DEBUG:PIL.Image:Importing ImtImagePlugin
+DEBUG:PIL.Image:Importing IptcImagePlugin
+DEBUG:PIL.Image:Importing JpegImagePlugin
+DEBUG:PIL.Image:Importing Jpeg2KImagePlugin
+DEBUG:PIL.Image:Importing McIdasImagePlugin
+DEBUG:PIL.Image:Importing MicImagePlugin
+DEBUG:PIL.Image:Image: failed to import MicImagePlugin: No module named 'olefile'
+DEBUG:PIL.Image:Importing MpegImagePlugin
+DEBUG:PIL.Image:Importing MpoImagePlugin
+DEBUG:PIL.Image:Importing MspImagePlugin
+DEBUG:PIL.Image:Importing PalmImagePlugin
+DEBUG:PIL.Image:Importing PcdImagePlugin
+DEBUG:PIL.Image:Importing PcxImagePlugin
+DEBUG:PIL.Image:Importing PdfImagePlugin
+DEBUG:PIL.Image:Importing PixarImagePlugin
+DEBUG:PIL.Image:Importing PngImagePlugin
+DEBUG:PIL.Image:Importing PpmImagePlugin
+DEBUG:PIL.Image:Importing PsdImagePlugin
+DEBUG:PIL.Image:Importing QoiImagePlugin
+DEBUG:PIL.Image:Importing SgiImagePlugin
+DEBUG:PIL.Image:Importing SpiderImagePlugin
+DEBUG:PIL.Image:Importing SunImagePlugin
+DEBUG:PIL.Image:Importing TgaImagePlugin
+DEBUG:PIL.Image:Importing TiffImagePlugin
+DEBUG:PIL.Image:Importing WebPImagePlugin
+DEBUG:PIL.Image:Importing WmfImagePlugin
+DEBUG:PIL.Image:Importing XbmImagePlugin
+DEBUG:PIL.Image:Importing XpmImagePlugin
+DEBUG:PIL.Image:Importing XVThumbImagePlugin
 
   8%|▊         | 1/12 [01:13<13:30, 73.69s/it]
  17%|█▋        | 2/12 [02:26<12:12, 73.23s/it]
@@ -72,6 +125,6 @@ DEBUG:PIL.PngImagePlugin:STREAM b'IDAT' 62 4096
 >>>>>>>>>>>>>>>>>>>>>>>>
 >>> Step 7: generating actions ...
 >>> Step 7: interacting with world model ...
->>>>>>>>>>>>>>>>>>>>>>>>
->>> Step 8: generating actions ...
->>> Step 8: interacting with world model ...
+>>>>>>>>>>>>>>>>>>>>>>>>
+>>> Step 8: generating actions ...
+>>> Step 8: interacting with world model ...

unitree_z1_stackbox/case4/psnr_result.json

@@ -1,5 +1,5 @@
 {
     "gt_video": "unitree_z1_stackbox/case4/unitree_z1_stackbox_case4.mp4",
     "pred_video": "unitree_z1_stackbox/case4/output/inference/35_full_fs4.mp4",
-    "psnr": 47.18724378194084
+    "psnr": 50.642542240144444
 }

unitree_z1_stackbox/case4/run_world_model_interaction.sh

@@ -20,6 +20,5 @@ dataset="unitree_z1_stackbox"
         --n_iter 12 \
         --timestep_spacing 'uniform_trailing' \
         --guidance_rescale 0.7 \
-        --perframe_ae \
-        --fast_policy_no_decode
+        --perframe_ae
 } 2>&1 | tee "${res_dir}/output.log"