fix(base): correct asserts for huggingface output

Author: AmitMY

.pre-commit-config.yaml

@@ -1,6 +1,5 @@
 repos:
   - repo: local
-    files: ^signwriting/
     hooks:
       - id: pylint
         name: pylint
@@ -9,7 +8,8 @@ repos:
         types: [python]
       - id: pytest
         name: pytest
-        entry: pytest signwriting_evaluation
+        entry: pytest .
         language: system
         types: [python]
+        pass_filenames: false
 

signwriting_evaluation/metrics/base.py

@@ -19,18 +19,21 @@ def score_max(self, hypothesis: str, references: list[str]) -> float:
 
     def validate_corpus_score_input(self, hypotheses: list[str], references: list[list[str]]):
         # This method is designed to avoid mistakes in the use of the corpus_score method
-        assert type(hypotheses) == list, "Hypotheses must be a list"
-        assert type(references) == list, "References must be a list"
+        assert isinstance(hypotheses,list), "Hypotheses must be a list"
+        assert isinstance(references, list), "References must be a list"
         if len(references) > 0:
             reference_type = type(references[0])
-            assert reference_type == list, f"References must be a list of lists (found list of {reference_type})"
+            assert reference_type in [list, tuple], \
+                f"References must be a list of lists or tuples (found list of {reference_type})"
 
         for reference in references:
             assert len(hypotheses) == len(reference), \
-                f"Hypotheses ({len(hypotheses)}) and reference ({len(references)}) must have the same number of instances"
+                (f"Hypotheses ({len(hypotheses)}) and reference ({len(reference)}) "
+                 f"must have the same number of instances (references is ({len(references)}))")
 
     def corpus_score(self, hypotheses: list[str], references: list[list[str]]) -> float:
         # Default implementation: average over sentence scores
+        # example: hypotheses=["hello"], references=[["hi"], ["hello"]]
         self.validate_corpus_score_input(hypotheses, references)
         transpose_references = list(zip(*references))
         return sum(self.score_max(h, r) for h, r in zip(hypotheses, transpose_references)) / len(hypotheses)

signwriting_evaluation/metrics/similarity.py

@@ -99,7 +99,6 @@ def error_rate(self, hyp: Sign, ref: Sign) -> float:
         return length_weight + mean_cost * (1 - length_weight)
 
     def score_single_sign(self, hypothesis: str, reference: str) -> float:
-        print("scoring", hypothesis, reference)
         # Calculate the evaluate score for a given hypothesis and ref.
         hyp = fsw_to_sign(hypothesis)
         ref = fsw_to_sign(reference)