SQuAD fixed to use loglikelihood API to calculate the probability of an unanswerable question.

lm_eval/tasks/squad.py

 import datasets
-from lm_eval.base import rf, f1_score, mean
+from math import exp
+from lm_eval.base import rf
+from lm_eval.metrics import f1_score, mean
 from . common import HFTask
 
 class SQuAD(HFTask):
@@ -48,7 +50,8 @@ class SQuAD(HFTask):
             part of the document for `doc`. 
         """
         continuation = rf.greedy_until(ctx, ['\n'])
-        return continuation
+        is_unanswerable = rf.loglikelihood(ctx, [' unanswerable'])
+        return continuation, is_unanswerable
     
     def process_results(self, doc, results):
         """Take a single document and the LM results and evaluates, returning a 
@@ -62,12 +65,12 @@ class SQuAD(HFTask):
         """
         squad_metric = datasets.load_metric("squad_v2")
 
-        continuation, = results
+        continuation, is_unanswerable = results
 
-        no_answer_probability = 0.0
-        if continuation.startswith(' unanswerable'):
-            no_answer_probability = 1.0
+        logprob_unanswerable, is_greedy = is_unanswerable
 
+        no_answer_probability = exp(logprob_unanswerable)
+        
         predictions = [{
             'id': doc['id'],
             'prediction_text': continuation,