README
Mobile-O Post-Training Data
Unified Multimodal Post-Training · ~105K Quadruplet Samples
📌 Overview
This dataset is used for Stage 3: Unified Multimodal Post-Training of Mobile-O, a unified multimodal model for on-device understanding and generation.
The goal of this stage is to jointly improve both image generation and visual understanding through a multi-task objective using quadruplet samples.
📊 Dataset Format
Each sample is a quadruplet consisting of:
| Field | Description |
|---|---|
| Generation Prompt | Text prompt for image generation |
| Image | Corresponding image |
| Question | Visual understanding question about the image |
| Answer | Ground-truth answer |
Total samples: ~105K
🏋️ Training Details
- Stage: 3 — Unified Multimodal Post-Training
- Trainable components: DiT + MCP + LLM (via LoRA) + Visual Encoder
- Frozen components: VAE only
🔗 Related Resources
| Resource | Link |
|---|---|
| 📄 Paper | arXiv |
| 💻 Code | GitHub |
| 🤗 Pre-Training Data | Mobile-O-Pre-Train |
| 🤗 SFT Data | Mobile-O-SFT |
| 🤗 Model (0.5B) | Mobile-O-0.5B |
| 🤗 Model (1.5B) | Mobile-O-1.5B |
📄 Citation
@article{shaker2026mobileo,
title={Mobile-O: Unified Multimodal Understanding and Generation on Mobile Device},
author={Shaker, Abdelrahman and Heakl, Ahmed and Muhammad, Jaseel and Thawkar, Ritesh and Thawakar, Omkar and Li, Senmao and Cholakkal, Hisham and Reid, Ian and Xing, Eric P. and Khan, Salman and Khan, Fahad Shahbaz},
journal={arXiv preprint arXiv:2602.20161},
year={2026}
}
⚖️ License
This dataset is released under CC BY-NC 4.0. For research purposes only.
Contributors
27 commits
Similar repos
README
Mobile-O Post-Training Data
Unified Multimodal Post-Training · ~105K Quadruplet Samples
📌 Overview
This dataset is used for Stage 3: Unified Multimodal Post-Training of Mobile-O, a unified multimodal model for on-device understanding and generation.
The goal of this stage is to jointly improve both image generation and visual understanding through a multi-task objective using quadruplet samples.
📊 Dataset Format
Each sample is a quadruplet consisting of:
| Field | Description |
|---|---|
| Generation Prompt | Text prompt for image generation |
| Image | Corresponding image |
| Question | Visual understanding question about the image |
| Answer | Ground-truth answer |
Total samples: ~105K
🏋️ Training Details
- Stage: 3 — Unified Multimodal Post-Training
- Trainable components: DiT + MCP + LLM (via LoRA) + Visual Encoder
- Frozen components: VAE only
🔗 Related Resources
| Resource | Link |
|---|---|
| 📄 Paper | arXiv |
| 💻 Code | GitHub |
| 🤗 Pre-Training Data | Mobile-O-Pre-Train |
| 🤗 SFT Data | Mobile-O-SFT |
| 🤗 Model (0.5B) | Mobile-O-0.5B |
| 🤗 Model (1.5B) | Mobile-O-1.5B |
📄 Citation
@article{shaker2026mobileo,
title={Mobile-O: Unified Multimodal Understanding and Generation on Mobile Device},
author={Shaker, Abdelrahman and Heakl, Ahmed and Muhammad, Jaseel and Thawkar, Ritesh and Thawakar, Omkar and Li, Senmao and Cholakkal, Hisham and Reid, Ian and Xing, Eric P. and Khan, Salman and Khan, Fahad Shahbaz},
journal={arXiv preprint arXiv:2602.20161},
year={2026}
}
⚖️ License
This dataset is released under CC BY-NC 4.0. For research purposes only.
Similar repos
Linked in READMEs
Contributors
27 commits