Text

app.py

@@ -89,7 +89,7 @@ def init_leaderboard(dataframe):
 demo = gr.Blocks(css=custom_css)
 with demo:
     gr.HTML(TITLE, elem_classes="abl_header")
-    gr.Markdown(INTRODUCTION_TEXT, elem_classes="abl_desc_text")
+    gr.HTML(INTRODUCTION_TEXT, elem_classes="abl_desc_text")
 
     with gr.Tabs(elem_classes="tab-buttons") as tabs:
         with gr.TabItem("🏅 Arabic LLM Leaderboard", elem_id="llm-benchmark-tab-table", id=0):

src/about.py

@@ -23,8 +23,12 @@ TITLE = """<div ><img class='abl_header_image' src='https://huggingface.co/space
 
 # What does your leaderboard evaluate?
 INTRODUCTION_TEXT = """
-<h1 style='width: 100%;text-align: center;' id="space-title">Arabic Board Benchmark (ABL). The first comprehensive Leaderboard for Arabic LLMs</h1>
-ABL is the official leaderboard of <a href='https://huggingface.co/datasets/silma-ai/arabic-broad-benchmark' target='_blank'>Arabic Board Benchmark (ABB)</a>. ABB is a compact, yet comprehensive benchmark aiming to evaluate Arabic LLMs from all angels. The benchmark consists of <b>450</b> high quality questions sampled from <b>63</b> Arabic benchmarking datasets, evaluating <b>22 categories</b> some of which are unique to Arabic language such as Diacritization and Dialects. Find more details in the about Tab.
+<h1 style='width: 100%;text-align: center;' id="space-title">Arabic Board Leaderboard (ABL) - The first comprehensive Leaderboard for Arabic LLMs</h1>
+ABL is the official Leaderboard of <a href='https://huggingface.co/datasets/silma-ai/arabic-broad-benchmark' target='_blank'>Arabic Board Benchmark (ABB)</a>. 
+With advanced features and innovative visualizations, we provide the community with a comprehensive view of the capabilities of Arabic models, showcasing their speed, diverse skills while also defending against benchmarking contamination. 
+The benchmark consists of <b>450</b> high quality questions sampled from <b>63</b> Arabic benchmarking datasets, evaluating <b>22 categories and skills</b>.
+Find more details in the about Tab. 
+
 
 """