Fix task metric type mismatch (#743)

* fixing task metric type

* fix style

Author: Abelgurung

src/lighteval/tasks/default_tasks.py

@@ -6727,7 +6727,14 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=128,
-    metric=[Metrics.exact_match, Metrics.quasi_exact_match, Metrics.f1_score, Metrics.rougeL, "bleu_1", "bleu_4"],
+    metric=[
+        Metrics.exact_match,
+        Metrics.quasi_exact_match,
+        Metrics.f1_score,
+        Metrics.rougeL,
+        Metrics.bleu_1,
+        Metrics.bleu_4,
+    ],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -7752,7 +7759,7 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=1,
-    metric=[Metrics.loglikelihood_acc, "loglikelihood_f1"],
+    metric=[Metrics.loglikelihood_acc, Metrics.loglikelihood_f1],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -7784,7 +7791,7 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=1,
-    metric=[Metrics.loglikelihood_acc, "loglikelihood_f1"],
+    metric=[Metrics.loglikelihood_acc, Metrics.loglikelihood_f1],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -10103,7 +10110,14 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=128,
-    metric=[Metrics.exact_match, Metrics.quasi_exact_match, Metrics.f1_score, Metrics.rougeL, "bleu_1", "bleu_4"],
+    metric=[
+        Metrics.exact_match,
+        Metrics.quasi_exact_match,
+        Metrics.f1_score,
+        Metrics.rougeL,
+        Metrics.bleu_1,
+        Metrics.bleu_4,
+    ],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -10119,7 +10133,14 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=128,
-    metric=[Metrics.exact_match, Metrics.quasi_exact_match, Metrics.f1_score, Metrics.rougeL, "bleu_1", "bleu_4"],
+    metric=[
+        Metrics.exact_match,
+        Metrics.quasi_exact_match,
+        Metrics.f1_score,
+        Metrics.rougeL,
+        Metrics.bleu_1,
+        Metrics.bleu_4,
+    ],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -10135,7 +10156,14 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=128,
-    metric=[Metrics.exact_match, Metrics.quasi_exact_match, Metrics.f1_score, Metrics.rougeL, "bleu_1", "bleu_4"],
+    metric=[
+        Metrics.exact_match,
+        Metrics.quasi_exact_match,
+        Metrics.f1_score,
+        Metrics.rougeL,
+        Metrics.bleu_1,
+        Metrics.bleu_4,
+    ],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -10173,7 +10201,14 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=512,
-    metric=[Metrics.exact_match, Metrics.quasi_exact_match, Metrics.f1_score, Metrics.rougeL, "bleu_1", "bleu_4"],
+    metric=[
+        Metrics.exact_match,
+        Metrics.quasi_exact_match,
+        Metrics.f1_score,
+        Metrics.rougeL,
+        Metrics.bleu_1,
+        Metrics.bleu_4,
+    ],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -13744,7 +13779,14 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=100,
-    metric=[Metrics.exact_match, Metrics.quasi_exact_match, Metrics.f1_score, Metrics.rougeL, "bleu_1", "bleu_4"],
+    metric=[
+        Metrics.exact_match,
+        Metrics.quasi_exact_match,
+        Metrics.f1_score,
+        Metrics.rougeL,
+        Metrics.bleu_1,
+        Metrics.bleu_4,
+    ],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,
@@ -15407,7 +15449,7 @@
     few_shots_split=None,
     few_shots_select=None,
     generation_size=1,
-    metric=[Metrics.loglikelihood_acc, "multi_f1_numeric"],
+    metric=[Metrics.loglikelihood_acc, Metrics.multi_f1_numeric],
     stop_sequence=["\n"],
     trust_dataset=True,
     version=0,