Xylolabs/omnivla-original-Tensix-Components

robotics model by Xylolabs

From the publisher

Model card & documentation

Source preview

Read the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.

Original OmniVLA operator components for Tenstorrent Tensix [](https://huggingface.co/NHirose/omnivla-original) [](#formats) [](https://huggingface.co/Xylolabs/omnivla-original-Tensix-Components/blob/main/LICENSE) [](https://huggingface.co/Xylolabs/omnivla-original-Tensix-Components/blob/main/loadcomponents.py) [](https://docs.tenstorrent.com) tenstorrent · tensix · robotics · navigation · safetensors · bfloat16 · float32 · model-components Selected inference operator components from Hirose's original OmniVLA navigation policy, packaged by Xylolabs with original tensor values and dtypes preserved. The archive requires separate policy integration. This export needs about 15.09 GB of device memory for its component weights alone, plus room for working buffers. The 38 tensor files cover DINOv2/SigLIP penultimate-feature extraction, patch inputs, multimodal/pose projection, embedding, 32 bidirectional language-prefix blocks, final norm and action regression head. Rotary tables cover 608 positions. The original policy uses two 224×224 images and local goal pose. Unused language-model/vision heads, unused final vision blocks, distance head and the separate LoRA adapter are excluded. This is a component archive, requiring an architecture adapter, rather than a complete runnable checkpoint. No new quantization is applied. Tensor content is…

Xylolabs/omnivla-original-Tensix-Components on Hugging Face A short preview, not the full document.

Read the full model card ↗ · Preview checked 2026-11-14T09:35:06.667Z

Inside the original model card — Document outline

Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.

Links from the model card

References supplied by the publisher, not independently verified endorsements. Check the destination before downloading files or entering credentials.

![License

Documentation belongs to its respective authors. Reported project/model license: other. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.

Model overview

Xylolabs/omnivla-original-Tensix-Components is a robotics model repository published by Xylolabs on Hugging Face. The source reports the pytorch library.

This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.

Model facts

Task: robotics
Library: pytorch
Recent downloads (30 days): 0
Cumulative likes: 0
Hugging Face trending score: 0
Reported safetensors parameters: Not reported
Architecture: Not reported
License: other
Access: Not gated by Hugging Face
Created: 2026-11-14T06:53:13.000Z
Last modified: 2026-11-14T06:53:40.000Z

Compatibility and lineage

Reported languages: en

Reported base models: NHirose/omnivla-original

Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.

Use and evaluate

Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.

No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.

Model card and usage instructions

Model files and configuration

Community discussion