Update evaluations/eval.py

evaluations/eval.py

@@ -14,8 +14,8 @@ from datasets import Audio, Dataset, load_dataset, load_metric
 from transformers import AutoFeatureExtractor, pipeline
 
 
-indic_normalizer_factory = IndicNormalizerFactory()
-indic_normalizer = indic_normalizer_factory.get_normalizer('hi')
+#indic_normalizer_factory = IndicNormalizerFactory()
+#indic_normalizer = indic_normalizer_factory.get_normalizer('hi')
 
 def log_results(result: Dataset, args: Dict[str, str]):
     """DO NOT CHANGE. This function computes and logs the result metrics."""
@@ -71,11 +71,11 @@ def normalize_text(text: str) -> str:
 
     return text
 
-def normalize_text_indic(text:str) -> str:
-    lang='hi'
-    normalized = indic_normalizer.normalize(text)
-    processed = ' '.join(trivial_tokenize(normalized, lang))
-    return processed
+#def normalize_text_indic(text:str) -> str:
+#    lang='hi'
+#    normalized = indic_normalizer.normalize(text)
+#    processed = ' '.join(trivial_tokenize(normalized, lang))
+#    return processed
 
 
 def main(args):
@@ -105,8 +105,8 @@ def main(args):
 
         batch["prediction"] = prediction["text"]
         batch["target"] = normalize_text(batch["sentence"])
-        batch["prediction"] = normalize_text_indic(batch["prediction"] )
-        batch["target"] = normalize_text_indic(batch["target"] )
+        #batch["prediction"] = normalize_text_indic(batch["prediction"] )
+        #batch["target"] = normalize_text_indic(batch["target"] )
         return batch
 
     # run inference on all examples