Add library_name, fix pipeline_tag

This PR improves the model card by ensuring:

- there's a proper `pipeline_tag`, ensuring the model can be found at https://huggingface.co/models?pipeline_tag=reinforcement-learning
- the proper library is added, enabling "how to use" in the top right.

README.md

@@ -1,6 +1,9 @@
 ---
 license: mit
+library_name: diffusers
+pipeline_tag: reinforcement-learning
 ---
+
 <div align="center">
   <div style="margin-bottom: 30px">  <!-- 减少底部间距 -->
     <div style="display: flex; flex-direction: column; align-items: center; gap: 8px">  <!-- 新增垂直布局容器 -->
@@ -10,15 +13,15 @@ license: mit
     </div>
     <h2 style="font-size: 32px; margin: 20px 0;">Skill Expansion and Composition in Parameter Space</h2>
     <h4 style="color: #666; margin-bottom: 25px;">International Conference on Learning Representation (ICLR), 2025</h4>
-    <p align="center" style="margin: 20px 0;">
+    <p align="center" style="margin: 30px 0;">
       <a href="https://arxiv.org/abs/2502.05932">
         <img src="https://img.shields.io/badge/arXiv-2502.05932-b31b1b.svg">
       </a>
-      <!-- &nbsp;&nbsp; -->
+      &nbsp;&nbsp;
       <a href="https://ltlhuuu.github.io/PSEC/">
         <img src="https://img.shields.io/badge/🌐_Project_Page-PSEC-blue.svg">
       </a>
-      <!-- &nbsp;&nbsp; -->
+      &nbsp;&nbsp;
       <a href="https://arxiv.org/pdf/2502.05932.pdf">
         <img src="https://img.shields.io/badge/📑_Paper-PSEC-green.svg">
       </a>
@@ -31,10 +34,10 @@ license: mit
     🔥 Official Implementation
   </p>
   <p style="font-size: 18px; max-width: 800px; margin: 0 auto;">
-          <b>PSEC</b> is a novel framework designed to:
+            <img src="assets/icon.svg" width="20"> <b>PSEC</b> is a novel framework designed to:
   </p>
 </div>
-<div align="center">
+<div align="left">
   <p style="font-size: 15px; font-weight: 600; margin-bottom: 20px;">
     🚀 <b>Facilitate</b> efficient and flexible skill expansion and composition <br>
      🔄 <b>Iteratively evolve</b> the agents' capabilities<br>
@@ -99,18 +102,18 @@ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:~/.mujoco/mujoco210/bin
 ```
 ## Run experiments
 ### Pretrain
-Pretrain the model with the following command. Meanwhile there are pre-trained models, you can download them from [here](https://drive.google.com/drive/folders/1lpcShmYoKVt4YMH66JBiA0MhYEV9aEYy?usp=sharing).
+Pretrain the model with the following command. Meanwhile there are pre-trained models, you can download them from [here](https://huggingface.co/LTL07/PSEC).
 ```python
 export XLA_PYTHON_CLIENT_PREALLOCATE=False
 CUDA_VISIBLE_DEVICES=0 python launcher/examples/train_pretrain.py --variant 0 --seed 0
 ```
 ### LoRA finetune
-Train the skill policies with LoRA to achieve skill expansion. Meanwhile there are pre-trained models, you can download them from [here](https://drive.google.com/drive/folders/1lpcShmYoKVt4YMH66JBiA0MhYEV9aEYy?usp=sharing).
+Train the skill policies with LoRA to achieve skill expansion. Meanwhile there are pre-trained models, you can download them from [here](https://huggingface.co/LTL07/PSEC).
 ```python
 CUDA_VISIBLE_DEVICES=0 python launcher/examples/train_lora_finetune.py --com_method 0 --model_cls 'LoRALearner' --variant 0 --seed 0
 ```
 ### Context-aware Composition
-Train the context-aware modular to adaptively leverage different skill knowledge to solve the tasks. You can download the pretrained model and datasets from [here](https://drive.google.com/drive/folders/1lpcShmYoKVt4YMH66JBiA0MhYEV9aEYy?usp=sharing). Then, run the following command,
+Train the context-aware modular to adaptively leverage different skill knowledge to solve the tasks. You can download the pretrained model and datasets from [here](https://huggingface.co/LTL07/PSEC). Then, run the following command,
 ```python
 CUDA_VISIBLE_DEVICES=0 python launcher/examples/train_lora_finetune.py --com_method 0 --model_cls 'LoRASLearner' --variant 0 --seed 0
 ```