Update model card

README.md

@@ -0,0 +1,110 @@
+---
+base_model:
+- ACE-Brain/ACE-Brain-0.5-8B
+library_name: transformers
+---
+
+<div align="center">
+  <img src="https://raw.githubusercontent.com/ACE-Brain-Team/ACE-Brain-0.5/main/assets/logo.png" width=600>
+</div>
+
+<br/>
+
+<div align="center" style="line-height: 1;">
+  |
+  <a href="https://huggingface.co/ACE-Brain/ACE-Brain-0.5-8B" target="_blank">🤗 HuggingFace</a>
+  &nbsp;|
+  <a href="https://raw.githubusercontent.com/ACE-Brain-Team/ACE-Brain-0.5/main/assets/ACE-Brain-0.5.pdf" target="_blank">📔 Technical Report</a>
+  &nbsp;|
+  <a href="https://github.com/ACE-Brain-Team/ACE-Brain-0.5" target="_blank">🤖 Github</a>
+  &nbsp;|
+  <br/>
+</div>
+
+## Overview
+
+**ACE-Brain-0.5** is a unified embodied foundation model for Physical Agentic AI. It extends ACE-Brain-0 from an understanding-centric spatial model into a closed-loop embodied model that can perceive the physical world, plan under goals, act through robot bodies, monitor execution progress, and improve from accumulated experience.
+
+ACE-Brain-0.5 organizes robot intelligence into five tightly coupled cognitive functions: **Spatial Perception**, **Decision Making**, **Embodied Interaction**, **Self Monitoring**, and **Self Improvement**. A single 8B backbone instantiates the core perception-planning-action-evaluation loop, supporting object and affordance grounding, 3D and egocentric spatial reasoning, long-horizon task planning, navigation and manipulation action generation, and progress estimation for verification and recovery.
+
+<div align="center">
+  <img src="https://raw.githubusercontent.com/ACE-Brain-Team/ACE-Brain-0.5/main/assets/teaser.png" width=800>
+</div>
+
+## Key Features
+
+- Unified embodied foundation model for Physical Agentic AI
+- Closed-loop cognitive model spanning Spatial Perception, Decision Making, Embodied Interaction, Self Monitoring, and Self Improvement
+- SSR+ training paradigm extending Scaffold-Specialize-Reconcile with a Reactivate stage
+- Unified coverage across spatial grounding, executable planning, navigation, manipulation, and progress estimation
+
+## Method & Architecture
+
+ACE-Brain-0.5 uses a shared embodied backbone to encode heterogeneous inputs and maintain a unified scene-and-task representation, while dedicated interfaces decode this shared state into spatial grounding, executable subgoal planning, navigation and manipulation actions, and progress-estimation signals. Training follows **SSR+**, which inherits the spatial scaffold from ACE-Brain-0, specializes domain capabilities, reconciles task vectors through model merging, and applies a lightweight Reactivate stage to align output conventions across grounding, navigation, manipulation, and progress estimation.
+
+<div align="center">
+  <img src="https://raw.githubusercontent.com/ACE-Brain-Team/ACE-Brain-0.5/main/assets/architecture.png" width=800>
+</div>
+
+## Capability Evaluation
+
+ACE-Brain-0.5 is evaluated as a unified embodied foundation model rather than as a collection of task-specific specialists. The goal is to verify whether a single model can preserve broad spatial understanding while extending to planning, action generation, execution monitoring, and self-improvement.
+
+| Capability | Summary |
+|------------|---------|
+| **Spatial Perception** | Preserves strong spatial reasoning while extending to embodied grounding and affordance understanding. |
+| **Decision Making** | Evaluates planning and decision reasoning under embodied and driving scenarios. |
+| **Embodied Interaction** | Supports executable navigation decisions and continuous manipulation control. |
+| **Self Monitoring** | Estimates task progress for execution assessment and recovery. |
+| **Self Improvement** | Uses rollout feedback to improve behavior beyond static imitation. |
+
+## Inference Example
+
+```python
+from transformers import Qwen3VLForConditionalGeneration, AutoProcessor
+
+model = Qwen3VLForConditionalGeneration.from_pretrained(
+    "ACE-Brain/ACE-Brain-0.5-8B",
+    dtype="auto",
+    device_map="auto",
+)
+
+processor = AutoProcessor.from_pretrained("ACE-Brain/ACE-Brain-0.5-8B")
+
+messages = [
+    {
+        "role": "user",
+        "content": [
+            {
+                "type": "image",
+                "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
+            },
+            {"type": "text", "text": "Describe this image."},
+        ],
+    }
+]
+
+inputs = processor.apply_chat_template(
+    messages,
+    tokenize=True,
+    add_generation_prompt=True,
+    return_dict=True,
+    return_tensors="pt",
+)
+inputs = inputs.to(model.device)
+
+generated_ids = model.generate(**inputs, max_new_tokens=128)
+generated_ids_trimmed = [
+    out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
+]
+output_text = processor.batch_decode(
+    generated_ids_trimmed,
+    skip_special_tokens=True,
+    clean_up_tokenization_spaces=False,
+)
+print(output_text)
+```
+
+## Citation
+
+If you find ACE-Brain-0.5 useful for your research and applications, please consider citing the technical report. BibTeX will be added once the public citation metadata is finalized.