Update app.py

app.py

@@ -1,15 +1,15 @@
 import streamlit as st
 import soundfile as sf
-# Use a pipeline as a high-level helper
-from transformers import pipeline
-import shutil
+from transformers import pipeline, Wav2Vec2ForSequenceClassification, Wav2Vec2Tokenizer
 
-# Clear the Hugging Face cache
-shutil.rmtree("./transformers_cache", ignore_errors=True)
+# Load the model and tokenizer
+model_name = "sami606713/emotion_classification"
+
+# Initialize the pipeline
 try:
-    classifier = pipeline("audio-classification", model="sami606713/emotion_classification")
+    classifier = pipeline("audio-classification", model=model_name, tokenizer=model_name)
 except Exception as e:
-    st.write(f"{e}")
+    st.write(f"Error loading model: {e}")
 
 # Title and description
 st.title("Audio Emotion Classification")
@@ -20,15 +20,18 @@ uploaded_file = st.file_uploader("Choose an audio file...", type=["wav", "mp3",
 
 if uploaded_file is not None:
     # Load the audio file
-    audio_input,sample_rate=sf.read(uploaded_file)
+    audio_input, sample_rate = sf.read(uploaded_file)
 
     # Display the audio player
     st.audio(uploaded_file)
 
     # Perform emotion classification
     st.write("Classifying...")
-    predictions = classifier(audio_input)
-
-    # Display the results
-    for prediction in predictions:
-        st.write(f"Emotion: {prediction['label']}, Score: {prediction['score']:.2f}")
+    try:
+        predictions = classifier(audio_input, sampling_rate=sample_rate)
+        
+        # Display the results
+        for prediction in predictions:
+            st.write(f"Emotion: {prediction['label']}, Score: {prediction['score']:.2f}")
+    except Exception as e:
+        st.write(f"Error during classification: {e}")