Remove arxiv tag from metadata, add pipeline tag

This PR cleans up the model card metadata:
- Removes the `arxiv` tag from YAML (arXiv IDs should only appear in the Markdown content as links, not in the metadata).
- Adds the `pipeline_tag: text-classification` for better discoverability and correct pipeline integration.

The Markdown content remains unchanged, preserving all existing links and descriptions.

README.md

@@ -1,14 +1,14 @@
 ---
-license: mit
 base_model: Dream-org/Dream-v0-Instruct-7B
-tags:
-  - arxiv:2609.35472
-  - dllm-prm-gap
-  - discrete-diffusion
-  - gsm8k
-  - process-reward-model
-  - dream-7b
 library_name: peft
+license: mit
+pipeline_tag: text-classification
+tags:
+- dllm-prm-gap
+- discrete-diffusion
+- gsm8k
+- process-reward-model
+- dream-7b
 ---
 
 <h1 style="text-align: left; font-size: 1.6em; margin-bottom: 0.75em;">
@@ -48,4 +48,4 @@ model, report = load_diffusion_prm(
 
 **Role:** C3B bidirectional last-token ORM-protocol control.
 
-Paper: [https://arxiv.org/abs/2609.35472](https://arxiv.org/abs/2609.35472).
+Paper: [https://arxiv.org/abs/2609.35472](https://arxiv.org/abs/2609.35472).