Add pipeline_tag and paper/GitHub links

This PR improves the model card by adding the missing `pipeline_tag: text-generation` to the YAML metadata and by adding direct links to the paper and GitHub code repository in the README. The existing metadata and content have been preserved.

README.md

@@ -1,20 +1,24 @@
 ---
-license: mit
 base_model: deepseek-ai/DeepSeek-R1-0528-Qwen3-8B
 datasets:
-  - DORAEMONG/PRO-STEP-PRM-Data
+- DORAEMONG/PRO-STEP-PRM-Data
 language:
-  - en
+- en
 library_name: peft
+license: mit
+pipeline_tag: text-generation
 tags:
-  - process-reward-model
-  - prm
-  - retrieval-augmented-generation
-  - lora
+- process-reward-model
+- prm
+- retrieval-augmented-generation
+- lora
 ---
 
 # PRO-STEP: Process Reward Model for Agentic RAG
 
+**Paper**: [PRO-STEP: Step-level Process Reward Optimization for Retrieval-Augmented Generation](https://huggingface.co/papers/2609.01658)  
+**Code**: [github.com/keemminnke/PRO-Step](https://github.com/keemminnke/PRO-Step)
+
 The supervised step-level **Process Reward Model (PRM)** for [PRO-STEP](https://huggingface.co/DORAEMONG/PRO-STEP-Policy-7B). Trained to evaluate agentic-RAG steps along six axes (entity grounding, search query quality, reasoning, answer specificity, recovery, overconfidence).
 
 - **Base model**: DeepSeek-R1-0528-Qwen3-8B
@@ -58,4 +62,4 @@ tokenizer = AutoTokenizer.from_pretrained("deepseek-ai/DeepSeek-R1-0528-Qwen3-8B
   author={...},
   year={2026}
 }
-```
+```