Delete apppp.py

apppp.py

@@ -1,68 +0,0 @@
-import tensorflow as tf
-import joblib
-import numpy as np
-import gradio as gr
-import librosa
-import soundfile as sf
-import os
-
-# Load model and label encoder
-model = tf.keras.models.load_model("animal_sound_cnn.keras")
-label_encoder = joblib.load("label_encoder.joblib")
-
-def preprocess_audio(audio_path, target_shape=(64, 64)):
-    """
-    Convert audio file to spectrogram with correct shape for model
-    """
-    try:
-        # 1. Load audio file
-        y, sr = librosa.load(audio_path, sr=None)
-        
-        # 2. Create mel spectrogram
-        S = librosa.feature.melspectrogram(y=y, sr=sr, n_mels=target_shape[0])
-        log_S = librosa.power_to_db(S, ref=np.max)
-        
-        # 3. Resize to target dimensions (FIXED THIS LINE)
-        if log_S.shape[1] < target_shape[1]:
-            pad_width = target_shape[1] - log_S.shape[1]
-            log_S = np.pad(log_S, ((0, 0), (0, pad_width)), mode='constant')
-        else:
-            log_S = log_S[:, :target_shape[1]]
-            
-        # 4. Add channel dimension and batch dimension
-        spectrogram = log_S[np.newaxis, ..., np.newaxis]
-        
-        return spectrogram.astype(np.float32)
-    
-    except Exception as e:
-        print(f"Preprocessing error: {str(e)}")
-        return None
-
-def predict(audio_path):
-    try:
-        # 1. Preprocess audio
-        spectrogram = preprocess_audio(audio_path)
-        if spectrogram is None:
-            return "Error processing audio"
-            
-        # 2. Check input shape matches model expectations
-        print(f"Input shape: {spectrogram.shape}")  # Debug log
-        
-        # 3. Predict
-        pred = model.predict(spectrogram)
-        animal = label_encoder.inverse_transform([np.argmax(pred)])[0]
-        
-        return animal
-    except Exception as e:
-        return f"Prediction error: {str(e)}"
-
-# Update requirements.txt to include:
-# librosa==0.10.1
-# soundfile==0.12.1
-
-gr.Interface(
-    fn=predict,
-    inputs=gr.Audio(type="filepath"),
-    outputs="label",
-    examples=["example1.wav", "example2.wav"] if os.path.exists("example1.wav") else None
-).launch()