Improve model card: Add `library_name`, update license, enhance content, and update paper links

README.md

@@ -1,34 +1,32 @@
 ---
+language:
+- en
+license: apache-2.0
+pipeline_tag: text-generation
 tags:
 - text-generation
 - agent
 - tool-use
 - long-context
-license: other
-language:
-- en
-pipeline_tag: text-generation
+library_name: transformers
 ---
 
 <div style="display: flex; justify-content: center; align-items: center; gap: 20px;">
-  <img src="assets/sii.jpg" alt="SII" width="100px">
-  <img src="assets/asi.png" alt="ASI" width="100px">
-
+  <img src="https://github.com/GAIR-NLP/LIMI/raw/main/assets/sii.jpg" alt="SII" width="100px">
+  <img src="https://github.com/GAIR-NLP/LIMI/raw/main/assets/asi.png" alt="ASI" width="100px">
 </div>
-<div align="center>  
-  
-
+<div align="center">
 <a href="https://github.com/GAIR-NLP/LIMI" target="_blank" style="margin: 2px;">
-    <img alt="Chat" src="assets/teaser.jpg" style="display: inline-block; vertical-align: middle;"/>
+    <img alt="Chat" src="https://github.com/GAIR-NLP/LIMI/raw/main/assets/teaser.jpg" style="display: inline-block; vertical-align: middle;"/>
 </a>
-
 </div>
+<img src="https://github.com/GAIR-NLP/LIMI/raw/main/assets/demos.png">
 
 # LIMI: Less is More for Agency
 
-[![arXiv](https://img.shields.io/badge/arXiv-2509.17567-b31b1b.svg)](https://arxiv.org/pdf/2509.17567)
+[![arXiv](https://img.shields.io/badge/arXiv-2509.17567-b31b1b.svg)](https://huggingface.co/papers/2509.17567)
 [![GitHub](https://img.shields.io/badge/GitHub-Repository-green)](https://github.com/GAIR-NLP/LIMI)
-[![Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Dataset-blue)](https://huggingface.co/datasets/GAIR/LIMI)
+[![Hugging Face Dataset](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Dataset-blue)](https://huggingface.co/datasets/GAIR/LIMI)
 
 ---
 To learn more about LIMI, feel free to explore our documentation and resources. Our release consists of the following sections:
@@ -45,6 +43,9 @@ LIMI is an agentic model fine‑tuned from [GLM‑4.5](https://huggingface.co/za
 - Long‑context trajectory with tokenizer‑filtered samples
 - OpenAI‑style `messages` with optional function/tool calls
 
+## Paper
+The model was introduced in the paper [LIMI: Less is More for Agency](https://huggingface.co/papers/2509.17567).
+
 ## Model Details
 
 - Base model: `zai-org/GLM-4.5`
@@ -63,18 +64,17 @@ Our models achieve state-of-the-art performance across multiple agentic evaluati
 | **LIMI-Air** | **35.4** | **34.3** | **33.1** | **34.3** |
 | **LIMI** | **71.7** | **74.2** | **74.6** | **73.5** |
 
-For detailed benchmark results, experimental setup, and comprehensive comparisons, please refer to our [paper](https://arxiv.org/pdf/2509.17567). 
+For detailed benchmark results, experimental setup, and comprehensive comparisons, please refer to our [paper](https://huggingface.co/papers/2509.17567). 
 
 ## Model Zoo
 
-Our LIMO model is available on Hugging Face 🤗:
+Our LIMI model is available on Hugging Face 🤗:
 
 | Model | Backbone | Size | Link |
 |---|---|---|---|
 | LIMI | [GLM‑4.5](https://huggingface.co/zai-org/GLM-4.5) | 353B | https://huggingface.co/GAIR/LIMI |
 | LIMI‑Air | [GLM‑4.5‑Air](https://huggingface.co/zai-org/GLM-4.5-Air) | 107B | https://huggingface.co/GAIR/LIMI-Air |
 
-
 ## Datasets
 
 We release our datasets through Hugging Face 🤗:
@@ -82,10 +82,9 @@ We release our datasets through Hugging Face 🤗:
 - Summary: curated agentic SFT data (OpenAI `messages`, optional `tools`, normalized tool‑call arguments); current release contains ~78 high‑quality samples.
 - Link: https://huggingface.co/datasets/GAIR/LIMI
 
-## Quick Start
+## Quick Start (Sample Usage)
 
-<details>
-<summary>Start with HF Transformers</summary>
+### Start with HF Transformers
 
 ```python
 from transformers import AutoModelForCausalLM, AutoTokenizer
@@ -113,10 +112,7 @@ out = model.generate(
 print(tok.decode(out[0][inputs['input_ids'].shape[1]:], skip_special_tokens=True))
 ```
 
-</details>
-
-<details>
-<summary>Start with VLLM</summary>
+### Start with VLLM
 
 ```python
 from vllm import LLM, SamplingParams
@@ -129,8 +125,6 @@ out = llm.generate(text, SamplingParams(temperature=0.6, max_tokens=4096, top_p=
 print(out[0].outputs[0].text)
 ```
 
-</details>
-
 ## Prompting
 
 - Messages follow OpenAI chat format; include a grounding system message when helpful.
@@ -169,5 +163,4 @@ print(out[0].outputs[0].text)
       primaryClass={cs.AI},
       url={https://arxiv.org/abs/2509.17567}, 
 }
-```
-
+```