Update app.py

app.py

@@ -47,6 +47,21 @@ def classify_emotion(audio):
  emotion_classifier = foreign_class(source="speechbrain/emotion-recognition-wav2vec2-IEMOCAP", pymodule_file="custom_interface.py", classname="CustomEncoderWav2vec2Classifier")
  out_prob, score, index, text_lab = emotion_classifier.classify_file(audio)
  return emo_dict[text_lab[0]]
+
+def slider_logic(slider):
+ if slider == 1:
+ theshold = .98
+ elif slider == 2:
+ threshold = .88
+ elif slider == 3:
+ threshold = .77
+ elif slider == 4:
+ threshold = .66
+ elif slider == 5:
+ threshold = .55
+ else:
+ threshold = []
+ return threshold
  
 # Create a Gradio interface with audio file and text inputs
 def classify_toxicity(audio_file, text_input, classify_anxiety, emo_class, explitive_selection, slider):
@@ -59,6 +74,7 @@ def classify_toxicity(audio_file, text_input, classify_anxiety, emo_class, expli
  print("emo_class ", emo_class, "explitive select", explitive_selection)
 
  ## SLIDER ##
+ threshold = slider_logic(slider)
  
  #------- explitive call ---------------
  
@@ -94,10 +110,12 @@ def classify_toxicity(audio_file, text_input, classify_anxiety, emo_class, expli
  print("keys ", classification_output.keys())
 
  # plot.update(x=classification_df["labels"], y=classification_df["scores"])
- 
+ if toxicity_score > threshold:
+ print("threshold exceeded!!")
  return toxicity_score, classification_output, transcribed_text
  # return f"Toxicity Score ({available_models[selected_model]}): {toxicity_score:.4f}"
  else: 
+ threshold = slider_logic(slider)
  model = whisper.load_model("large")
  # model = model_cache[model_name]
  # class_names = classify_anxiety.split(",")