test_2

tasks/audio.py

@@ -20,7 +20,6 @@ DESCRIPTION = "Random Baseline"
 ROUTE = "/audio"
 def create_spec(dataset):
     spectograms = []
-    y = []
     
     for d in enumerate(dataset):  
         audio_sample = d["audio"]["array"] if d["audio"]["sampling_rate"] == 12000 else librosa.resample(
@@ -51,9 +50,8 @@ def create_spec(dataset):
         mel_db = librosa.power_to_db(mel, ref=np.max)  
         mel_db_normalized = (mel_db - mel_db.mean()) / (mel_db.std() + 1e-8)
         spectograms.append(np.float16(mel_db_normalized).T.flatten())
-        y.append(d["label"])
         
-    return np.stack(spectograms),y
+    return np.stack(spectograms)
 
 
 @router.post(ROUTE, tags=["Audio Task"],
@@ -86,9 +84,9 @@ async def evaluate_audio(request: AudioEvaluationRequest):
     tracker.start()
     tracker.start_task("inference")
     
-    test_spec, y_test = create_spec(test_dataset)
+    test_spec = create_spec(test_dataset)
     H = np.load("array_file.npy")
-    W_val = np.dot(test_spec, H)
+    W_test = np.dot(test_spec, H)
     model_filename = 'model.joblib'
     model = joblib.load(model_filename)
     #--------------------------------------------------------------------------------------------
@@ -98,7 +96,7 @@ async def evaluate_audio(request: AudioEvaluationRequest):
     
     # Make random predictions (placeholder for actual model inference)
     true_labels = test_dataset["label"]
-    predictions = model.predict(W_val)
+    predictions = model.predict(W_test)
     #--------------------------------------------------------------------------------------------
     # YOUR MODEL INFERENCE STOPS HERE
     #--------------------------------------------------------------------------------------------