[MODIFY] Cross-evaluation framework column names

src/about.py

@@ -64,10 +64,10 @@ class MedicalSummarizationColumn:
     col_name: str
 
 class MedicalSummarizationColumns(Enum):
-    medical_summarization_column0 = MedicalSummarizationColumn("brief", "score", "Brief Score")
-    medical_summarization_column1 = MedicalSummarizationColumn("coverage", "score", "Coverage Score")
-    medical_summarization_column2 = MedicalSummarizationColumn("conform", "score", "Conform Score")
-    medical_summarization_column3 = MedicalSummarizationColumn("fact", "score", "Fact Score")
+    medical_summarization_column0 = MedicalSummarizationColumn("coverage", "score", "Coverage")
+    medical_summarization_column1 = MedicalSummarizationColumn("conform", "score", "Conformity")
+    medical_summarization_column2 = MedicalSummarizationColumn("fact", "score", "Consistency")
+    medical_summarization_column3 = MedicalSummarizationColumn("brief", "score", "Conciseness")
 
 @dataclass
 class ACIColumn:
@@ -76,10 +76,10 @@ class ACIColumn:
     col_name: str
 
 class ACIColumns(Enum):
-    aci_column0 = ACIColumn("brief", "score", "Brief Score") 
-    aci_column1 = ACIColumn("coverage", "score", "Coverage Score") 
-    aci_column2 = ACIColumn("conform", "score", "Conform Score") 
-    aci_column3 = ACIColumn("fact", "score", "Fact Score") 
+    aci_column0 = ACIColumn("coverage", "score", "Coverage") 
+    aci_column1 = ACIColumn("conform", "score", "Conformity") 
+    aci_column2 = ACIColumn("fact", "score", "Consistency") 
+    aci_column3 = ACIColumn("brief", "score", "Conciseness") 
 
 @dataclass
 class SOAPColumn:
@@ -88,10 +88,10 @@ class SOAPColumn:
     col_name: str
 
 class SOAPColumns(Enum):
-    soap_column0 = SOAPColumn("brief", "score", "Brief Score") 
-    soap_column1 = SOAPColumn("coverage", "score", "Coverage Score") 
-    soap_column2 = SOAPColumn("conform", "score", "Conform Score") 
-    soap_column3 = SOAPColumn("fact", "score", "Fact Score") 
+    soap_column0 = SOAPColumn("coverage", "score", "Coverage") 
+    soap_column1 = SOAPColumn("conform", "score", "Conformity") 
+    soap_column2 = SOAPColumn("fact", "score", "Consistency") 
+    soap_column3 = SOAPColumn("brief", "score", "Conciseness") 
 
 NUM_FEWSHOT = 0  # Change with your few shot
 # ---------------------------------------------------

src/populate.py

@@ -25,11 +25,11 @@ def get_leaderboard_df(results_path: str, requests_path: str, cols: list, benchm
     elif subset == "open_ended":
         df = df.sort_values(by=["ELO"], ascending=False)
     elif subset == "medical_summarization":
-        df = df.sort_values(by=["Coverage Score"], ascending=False)
+        df = df.sort_values(by=["Coverage"], ascending=False)
     elif subset == "aci":
-        df = df.sort_values(by=["Coverage Score"], ascending=False)
+        df = df.sort_values(by=["Coverage"], ascending=False)
     elif subset == "soap":
-        df = df.sort_values(by=["Coverage Score"], ascending=False)
+        df = df.sort_values(by=["Coverage"], ascending=False)
     cols = list(set(df.columns).intersection(set(cols)))
     df = df[cols].round(decimals=2)
     # filter out if any of the benchmarks have not been produced