fixed bugs

utils.py

@@ -15,7 +15,7 @@ TASKS_V1 = ["V1-Overall", "I-CLS", "I-QA", "I-RET", "I-VG"]
 TASKS_V2 = ["V2-Overall", "V-CLS", "V-QA", "V-RET", "V-MRET", "VisDoc"]
 COLUMN_NAMES = BASE_COLS + TASKS_V1 + TASKS_V2
 
-DATA_TITLE_TYPE = ['number', 'markdown', 'str', 'markdown'] + 
+DATA_TITLE_TYPE = ['number', 'markdown', 'str', 'markdown'] + \
                     ['number'] * (len(TASKS_V1) + len(TASKS_V2))
 
 LEADERBOARD_INTRODUCTION = """
@@ -68,6 +68,7 @@ SUBMIT_INTRODUCTION = """# Submit on MMEB Leaderboard Introduction
         "I-VG": 50.0
     }, 
 ]
+```
 
 ### **TO SUBMIT V2 ONLY**
 ```json
@@ -77,7 +78,7 @@ SUBMIT_INTRODUCTION = """# Submit on MMEB Leaderboard Introduction
         <Optional>"URL": "<Model URL>",
         "Model Size(B)": 1000,
         "Data Source": Self-Reported,
-        "V1-Overall": 50.0,
+        "V2-Overall": 50.0,
         "V-CLS": 50.0,
         "V-QA": 50.0,
         "V-RET": 50.0,
@@ -86,8 +87,8 @@ SUBMIT_INTRODUCTION = """# Submit on MMEB Leaderboard Introduction
     }, 
 ]
 ```
-You are also welcome to submit both versions by including all the fields above! :)
-You may refer to the Github page for instructions about evaluating your model. 
+You are also welcome to submit both versions by including all the fields above! :) \n
+You may refer to the Github page for instructions about evaluating your model. \n
 Github link: https://github.com/TIGER-AI-Lab/VLM2Vec. \n
 Please send us an email at [email protected], attaching the JSON file. We will review your submission and update the leaderboard accordingly.
 """
@@ -104,21 +105,21 @@ def create_hyperlinked_names(df):
     df = df.apply(add_link_to_model_name, axis=1)
     return df
 
-def fetch_data(file: str) -> pd.DataFrame:
-    # fetch the leaderboard data from remote
-    if file is None:
-        raise ValueError("URL Not Provided")
-    url = f"https://huggingface.co/spaces/TIGER-Lab/MMEB/resolve/main/{file}"
-    print(f"Fetching data from {url}")
-    response = requests.get(url)
-    if response.status_code != 200:
-        raise requests.HTTPError(f"Failed to fetch data: HTTP status code {response.status_code}")
-    return pd.read_json(io.StringIO(response.text), orient='records', lines=True)
+# def fetch_data(file: str) -> pd.DataFrame:
+#     # fetch the leaderboard data from remote
+#     if file is None:
+#         raise ValueError("URL Not Provided")
+#     url = f"https://huggingface.co/spaces/TIGER-Lab/MMEB/resolve/main/{file}"
+#     print(f"Fetching data from {url}")
+#     response = requests.get(url)
+#     if response.status_code != 200:
+#         raise requests.HTTPError(f"Failed to fetch data: HTTP status code {response.status_code}")
+#     return pd.read_json(io.StringIO(response.text), orient='records', lines=True)
 
 def get_df(file="results.jsonl"):
-    df = fetch_data(file)
+    df = pd.read_json(file, orient='records', lines=True)
     df['Model Size(B)'] = df['Model Size(B)'].apply(process_model_size)
-    df = df.sort_values(by=['Overall'], ascending=False)
+    df = df.sort_values(by=['V1-Overall'], ascending=False)
     df = create_hyperlinked_names(df)
     df['Rank'] = range(1, len(df) + 1)
     return df