groundingpi / GroundAnything

English | 简体中文 
 Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed 
 Demo Video · Quick Start · Documentation · Citation 
## 📰 News
- **2026-10-14:** GroundAnything and GroundAnything-VLM model weights are now available on ModelScope.
- **2026-10-14:** GroundAnything received strong community recognition, with **74 upvotes** on Hugging Face Daily Papers.
- **2026-10-04:** Released the project webpage.
- **2026-10-03:** Released the source code, deployment and batch-annotation guides, and full-suite evaluation workflows.
- **2026-10-01:** We released the GroundAnything and GroundAnything-VLM model weights on Hugging Face.
- **2026-09-30:** The GroundAnything paper is available on arXiv.
## 🧭 Contents
Highlights · Demo · Models · Installation · Deployment Options · Quick Start · vLLM Deployment · Batch Annotation · Tasks and Output Format · Method and Inference Infrastructure · Evaluation · Training · Results · Documentation · License · Citation · Acknowledgement
## ✨ Highlights
- **Unlocking diffusion for broad visual grounding.** We introduce GroundAnything, a 4B foundation model that unifies diverse grounding tasks with precise localization and parallel decoding, trained through grounding pretraining, AR-to-diffusion conversion with joint objectives, supervised fine-tuning, and GRPO-based reinforcement learning.
- **State-of-the-art grounding at 4B scale.** Across 30 benchmarks, GroundAnything surpasses the prior overall state of the art among similarly sized AR models and outperforms the larger Qwen3.7-Max. GroundAnything-VLM establishes a new overall state of the art at this scale and remains competitive with GPT-6 Astra.
- **Systematic acceleration studies.** We analyze decoding strategies in detail, compare with MTP-based generation, and evaluate progressive infrastructure optimizations for practical deployment.
## 🎬 Demo
▶ Watch the demo
▶ Parallel Decoding
## 🤗 Models
| Checkpoint | Generation | Download | Evaluation mode |
|:---|:---|:---|:---|
| **GroundAnything** | Entropy-guided diffusion; optional self-speculation | DLM weights | **GAM** |
| **GroundAnything-VLM** | Autoregressive generation | VLM weights | **GAM** |
GroundAnything's main results use entropy-guided decoding. GroundAnything-VLM is a separate autoregressive checkpoint. Both use the **GAM** evaluation mode.
## 🛠️ Installation
The serving workflows below target **Linux x86_64 and Python 3.12**. The serving setup installs the bundled custom **SGLang** engine and its pinned dependencies. The supplied serving profile uses Torch **2.9.1**, Transformers **5.5.4**, Triton **3.5.1**, sgl-kernel **0.3.20**, and FlashInfer **0.5.3**.
 [code] 
Already have a source checkout? Start with `cd GroundAnything`. Run subsequent commands from the repository root.
`requirements.txt` installs the lightweight HTTP client, visualization tools, and setup dependencies. Serving, training, and evaluation each use their own environment; `pip install -r requirements.txt` alo