import joblib
from sklearn.preprocessing import LabelEncoder

# Assuming you have the dataset path (same as in your training code)
#dataset_path = "path/to/your/dataset"  # Update this

# Initialize label encoder (same as in VoiceDataset)
label_encoder = LabelEncoder()

# Extract labels from dataset folders
labels = []
for user_folder in os.listdir(dataset_path):
    if os.path.isdir(os.path.join(dataset_path, user_folder)):
        labels.append(user_folder)

# Fit the label encoder
label_encoder.fit(labels)

# Save to file
joblib.dump(label_encoder, "label_encoder.joblib")
print(f"Label encoder saved with classes: {label_encoder.classes_}")