---
license: apache-2.0
base_model:
- apple/aimv2-large-patch14-448
pipeline_tag: zero-shot-object-detection
tags:
- t5gemma
- vision
---

# TMoon

- 3D positional embedding (as in the [text encoder replacement](https://huggingface.co/sugarquark/sd15-text-encoder-t5g-2b-ul2-it), [object detection](https://huggingface.co/twodgirl/rotated-position-in-latent-space-fashion-object-detection) and others)
- Supports multiple image aspect ratios
- Contrastive prediction model
- AIMv2 vision model
- [T5Gemma](https://huggingface.co/google/t5gemma-2b-2b-ul2-it) text encoder
- [Moondream2](https://huggingface.co/vikhyatk/moondream2/tree/2025-06-21) [dataset](https://huggingface.co/datasets/sugarquark/moonpixs)

![](images/horizon.jpg)