Add Vaani Multilingual langauge detection and gender classification task

mteb/models/clap_models.py

@@ -62,6 +62,8 @@ def _handle_batch(
                     if "array" in item:
                         audio = item["array"]
                         # Convert to torch tensor and ensure float32
+                        if isinstance(audio, list):
+                            audio = np.array(audio)
                         audio = (
                             torch.from_numpy(audio).float()
                             if isinstance(audio, np.ndarray)

mteb/models/wav2vec2_models.py

@@ -112,6 +112,8 @@ def _handle_batch(
                 if isinstance(item, dict):
                     if "array" in item:
                         audio = item["array"]
+                        if isinstance(audio, list):
+                            audio = np.array(audio)
                         audio = (
                             torch.from_numpy(audio).float()
                             if isinstance(audio, np.ndarray)

mteb/tasks/Audio/AudioClassification/__init__.py

@@ -5,4 +5,5 @@
 from .eng.GTZANGenre import *
 from .eng.GunshotTriangulation import *
 from .eng.NSynth import *
+from .eng.VaaniGenderClassification import *
 from .eng.VoxLingua107Top10 import *