Cosmos predict 2.5

fcj · April 11, 2026, 5:47am

How to run cosmos predict 2.5 on dgx spark?

Is there an example or a docker container for this?

Thanks,

aniculescu · April 13, 2026, 8:13pm

Have you tried following the setup instructions in the repo for Blackwell? cosmos-predict2.5/docs/setup.md at main · nvidia-cosmos/cosmos-predict2.5 · GitHub

fcj · April 13, 2026, 8:36pm

I thought there is an issue of

github.com/nvidia-cosmos/cosmos-predict2.5

NVIDIA Cosmos (Predict2.5) on DGX Spark (ARM / aarch64)

opened 06:13PM - 06 Feb 26 UTC

amirtd

I'm not able to run NVIDIA Cosmos (Predict2.5 / text2world) on DGX Spark (ARM / …aarch64)? I’m hitting dependency blockers around Transformer Engine / flash-attn and some configs that seem x86_64-centric. Curious if anyone has a confirmed working setup, workarounds, or if Cosmos is currently x86-only in practice. Environment - Hardware: NVIDIA DGX Spark - Architecture: aarch64 (ARM) - OS: Ubuntu (Spark image) - Kernel: 6.14.0-1015-nvidia - GPU: NVIDIA GB10 - Driver: 580.126.09 - CUDA: 13.0 - Python: 3.10 (Conda) - Cosmos Version: cosmos-predict2.5 ### Key Issues Encountered **1. Transformer Engine Is a Hard Dependency (and Not Available on ARM)** Multiple Cosmos modules unconditionally import transformer_engine, including: - minimal_v4_dit.py - wan2pt1.py - fused_adam_dtensor.py - optimizer and attention layers Example failure: > ModuleNotFoundError: No module named 'transformer_engine' Even after: - Rewriting attention layers - Introducing fallback implementations - Removing explicit class inheritance from DotProductAttention …the import graph still eventually requires transformer_engine. **Root problem:** transformer-engine wheels are not available for aarch64, and building from source is not supported on DGX Spark. **2. flash-attn / flash-attn-2 Is x86_64-Centric** Cosmos enforces availability checks such as: > assert is_flash_attn_2_available() On DGX Spark: - flash-attn is not available - flash-attn-2 cannot be installed - CUDA + ARM + GB10 has no supported build This causes early termination even when the feature is not strictly required for inference. **3. Config System Assumes Missing Experiment Modules** Text2world config attempts to auto-import: > cosmos_predict2._src.predict2.configs.text2world.experiment But the directory does not exist in the repo, leading to: > ModuleNotFoundError: No module named 'cosmos_predict2._src.predict2.configs.text2world.experiment' This suggests: - Incomplete packaging - Internal-only config assumptions - Or missing repo components **4. Inference Path Is Hardwired to Video2World** Even with: - --inference-type=text2world - Environment overrides - Patched model_loader.py The system still instantiates: > Video2WorldInference(...) This drags in: - Vision-language models - Qwen-VL - Transformer Engine - Flash attention paths There is no clean text-only inference path that avoids these dependencies. **5. Cascading Failures from Partial Patching** Attempts to patch around missing dependencies led to secondary failures: - NoneType takes no arguments (class inheritance on disabled components) - NameError: _SACConfig not defined - SyntaxError: return outside function (after fallback edits) This indicates the codebase is not designed for partial feature removal. What Would Be Needed to Support DGX Spark 1. Optional Transformer Engine - Clean fallbacks - No hard imports - No class inheritance from TE types 2. flash-attn Optionality - Runtime feature flags - Safe PyTorch attention fallback 3. True Text-Only Inference Path - No Video2World initialization - No vision or VLM imports 4. ARM-Compatible Build Artifacts - TE or equivalent kernels for aarch64 - Or official statement of non-support

Related context from Transformer Engine: support on DGX Spark (SM120) is still being actively brought up (see NVIDIA/TransformerEngine#2668)

Has this issue been resolved?

Topic		Replies	Views
Cosmos predict 2.5 on dgx spark Tutorials & Guides cosmos	4	75	May 11, 2026
Nvidia Cosmos running on Jetson Jetson Projects jetson , cosmos	13	935	January 13, 2025
Cosmos Reason 2 on Drive AGX Thor DRIVE AGX Thor General	6	74	April 30, 2026
What is difference between the models in NVIDIA/cosmos and nvidia-cosmos/cosmos-predict1? General Discussion cosmos	3	204	March 31, 2025
NVIDIA Cosmos on AGX Orin issue Jetson AGX Orin cosmos	10	349	April 1, 2025
Has anyone been able to get Ostris' AI Toolkit running on DGX Spark? DGX Spark / GB10	22	2964	December 19, 2025
Unsloth Studio, GUI to train models DGX Spark / GB10	12	1310	May 4, 2026
Issues running cosmos-reason1 on Jetson AGX orin Jetson AGX Orin containers , cosmos	10	311	November 10, 2025
Anyone got nanochat training working on the DGX spark? DGX Spark / GB10	10	1603	November 21, 2025
[Guide] Running Cosmos Reason2 on AGX Orin + JetPack 6.2.2 from Ubuntu 24 — FP8/SM 8.7 fix included Jetson Projects edgeai , cosmos , physical-ai	2	104	March 25, 2026

Cosmos predict 2.5

Related topics