eggie5-adyen commited on
Commit
727eb6f
·
1 Parent(s): f477fda

added Validated tab

Browse files
Files changed (2) hide show
  1. app.py +24 -6
  2. dabstep_benchmark/leaderboard.py +5 -2
app.py CHANGED
@@ -18,13 +18,31 @@ if __name__ == "__main__":
18
  with demo:
19
  gr.Markdown(TITLE)
20
  gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text")
 
 
 
21
 
22
- leaderboard_table = gr.components.Dataframe(
23
- value=generate_leaderboard_df(),
24
- datatype=["markdown", "str", "str", "str", "markdown", "str", "str", "str"],
25
- interactive=False,
26
- column_widths=["20%"],
27
- wrap=True,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
  )
29
  # create a Gradio event listener that runs when the page is loaded to populate the dataframe
30
  demo.load(lambda: generate_leaderboard_df(), None, leaderboard_table)
 
18
  with demo:
19
  gr.Markdown(TITLE)
20
  gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text")
21
+
22
+ # Generate leaderboard data once
23
+ leaderboard_df = generate_leaderboard_df()
24
 
25
+ # Filter validated and unvalidated
26
+ validated = leaderboard_df[leaderboard_df["validated"] == True].drop(columns=["validated"]).copy()
27
+ unvalidated = leaderboard_df[leaderboard_df["validated"] == False].drop(columns=["validated"]).copy()
28
+
29
+
30
+ with gr.Tab("Validated"):
31
+ leaderboard_table = gr.components.Dataframe(
32
+ value=validated,
33
+ datatype=["markdown", "str", "str", "str", "markdown", "str", "str", "str"],
34
+ interactive=False,
35
+ column_widths=["20%"],
36
+ wrap=True,
37
+ )
38
+
39
+ with gr.Tab("Unvalidated"):
40
+ leaderboard_table = gr.components.Dataframe(
41
+ value=unvalidated,
42
+ datatype=["markdown", "str", "str", "str", "markdown", "str", "str", "str"],
43
+ interactive=False,
44
+ column_widths=["20%"],
45
+ wrap=True,
46
  )
47
  # create a Gradio event listener that runs when the page is loaded to populate the dataframe
48
  demo.load(lambda: generate_leaderboard_df(), None, leaderboard_table)
dabstep_benchmark/leaderboard.py CHANGED
@@ -154,6 +154,7 @@ def process_submission(
154
  submission_df["organisation"] = f"{organisation} | user {profile.username}"
155
  submission_df["repo_url"] = repo_url
156
  submission_df["date"] = datetime.date.today().strftime("%d-%m-%Y")
 
157
 
158
  # add empty reasoning trace if one is not provided to not break schema of datasets
159
  if "reasoning_trace" not in submission_df.columns:
@@ -242,7 +243,8 @@ def generate_leaderboard_df() -> pd.DataFrame:
242
  "model_family",
243
  "organisation",
244
  "repo_url",
245
- "date"
 
246
  ]
247
  ]
248
  )
@@ -288,7 +290,8 @@ def generate_leaderboard_df() -> pd.DataFrame:
288
  "organisation": "Organization",
289
  "repo_url": "Repo URL",
290
  "model_family": "Model Family",
291
- "date": "Date"
 
292
  }
293
  col_order = [new_col_name for new_col_name in col_map.values()]
294
  leaderboard_df.rename(columns=col_map, inplace=True)
 
154
  submission_df["organisation"] = f"{organisation} | user {profile.username}"
155
  submission_df["repo_url"] = repo_url
156
  submission_df["date"] = datetime.date.today().strftime("%d-%m-%Y")
157
+ submission_df["validated"] = False #unvalidated by default
158
 
159
  # add empty reasoning trace if one is not provided to not break schema of datasets
160
  if "reasoning_trace" not in submission_df.columns:
 
243
  "model_family",
244
  "organisation",
245
  "repo_url",
246
+ "date",
247
+ "validated"
248
  ]
249
  ]
250
  )
 
290
  "organisation": "Organization",
291
  "repo_url": "Repo URL",
292
  "model_family": "Model Family",
293
+ "date": "Date",
294
+ "validated": "validated"
295
  }
296
  col_order = [new_col_name for new_col_name in col_map.values()]
297
  leaderboard_df.rename(columns=col_map, inplace=True)