Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Tarsier -- a family of large-scale video-language models, which is designed to generate high-quality video descriptions , together with good capability of general video understanding.
| Date | Stars |
|---|---|
| 2026-07-31 | 548 |
| 2026-08-03 | 548 |
| 2026-08-06 | 548 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<div align="center">
<h2><a href="https://github.com/bytedance/tarsier/tree/tarsier2">Tarsier2: Advancing Large Vision-Language Models from Detailed Video Description to Comprehensive Video Understanding</a></h2>
Liping Yuan*, Jiawei Wang*, Haomiao Sun*, Yuchen Zhang*, Yuan Lin†
ByteDance Research
*Equally contributed. †Corresponding author.
</div>
<!-- [](https://arxiv.org/abs/2311.17005) -->
[](https://arxiv.org/abs/2501.07888)(Tarsier2)
[](https://arxiv.org/abs/2407.00634)(Tarsier)
[](https://huggingface.co/spaces/omni-research/Tarsier2-7b)
[](https://huggingface.co/omni-research/Tarsier-34b)
[](https://huggingface.co/datasets/omni-research/DREAM-1K)
[](https://paperswithcode.com/sota/video-question-answering-on-mvbench?p=tarsier-recipes-for-training-and-evaluating-1)
[](https://paperswithcode.com/sota/zero-shot-video-question-answer-on-next-qa?p=tarsier-recipes-for-training-and-evaluating-1)
[](https://paperswithcode.com/sota/zero-shot-video-question-answer-on-egoschema-1?p=tarsier-recipes-for-training-and-evaluating-1)
[](https://paperswithcode.com/sota/zeroshot-video-question-answer-on-msvd-qa?p=tarsier-recipes-for-training-and-evaluating-1)
[](https://paperswithcode.com/sota/zeroshot-video-question-answer-on-tgif-qa?p=tarsier-recipes-for-training-and-evaluating-1)
[](https://paperswithcode.com/sota/zeroshot-video-question-answer-on-activitynet?p=tarsier-recipes-for-training-and-evaluating-1)
[](https://paperswithcode.com/sota/zeroshot-video-question-answer-on-msrvtt-qa?p=tarsier-recipes-for-training-and-evaluating-1)
<div align="center">
<a href="https://github.com/bytedance/tarsier">
<img src="assets/figures/tarsier_logo.jpg" width = "60%">
</a>
</div>
# Release Notes
- [2025/04/25] 🔥🚀✨ **[Tarsier2-7B(-0115)](https://huggingface.co/omni-research/Tarsier2-7b-0115)** is released! **It is the exact model we tested in [Tarsier2 Technical Report](https://arxiv.org/abs/2501.07888).** As it was only exposed to video caption tasks in post-training (both SFT and DPO), you may find it difficult to follow instructions at times. If you are testing it on multi-choice tasks, we recommend you format your prompt following the [Data Prepare](https://github.com/bytedance/tarsier?tab=readme-ov-file#data-prepare) part. Also, we are plaExcerpt of 26,338 characters
Read on GitHub
43
汪志鹏 · Rakuten Asia Pte Ltd · Singapore
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:6f2d2ebf7b9037e3, llm:Repository description: 'Tarsier -- a family of large-scale video-language models... generate high-quality video descriptions, together with good capability of general video understanding.' Topics: research. Language: Python.
matched fp:6f2d2ebf7b9037e3, llm:Repository description: 'Tarsier -- a family of large-scale video-language models... generate high-quality video descriptions, together with good capability of general video understanding.' Topics: research. Language: Python.
matched fp:6f2d2ebf7b9037e3, llm:Repository description: 'Tarsier -- a family of large-scale video-language models... generate high-quality video descriptions, together with good capability of general video understanding.' Topics: research. Language: Python.
matched fp:6f2d2ebf7b9037e3, llm:Repository description: 'Tarsier -- a family of large-scale video-language models... generate high-quality video descriptions, together with good capability of general video understanding.' Topics: research. Language: Python.