app.py

import os
import torch
import torchaudio
import numpy as np
import gradio as gr
from huggingface_hub import hf_hub_download
model_path = hf_hub_download(
    repo_id="DennisHung/Pre-trained_AudioMAE_weights",
    filename="pretrained.pth",
    local_dir="./",                                
    local_dir_use_symlinks=False 
)

model_path = hf_hub_download(
    repo_id="DennisHung/Pre-trained_AudioMAE_weights",
    filename="pytorch_model.bin",
    local_dir="./",                                
    local_dir_use_symlinks=False 
)

from pipeline.morph_pipeline_successed_ver1 import AudioLDM2MorphPipeline 
# Initialize AudioLDM2 Pipeline
pipeline = AudioLDM2MorphPipeline.from_pretrained("cvssp/audioldm2-large", torch_dtype=torch.float32)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
pipeline.to(device)

# Audio morphing function
def morph_audio(audio_file1, audio_file2, prompt1, prompt2, negative_prompt1="Low quality", negative_prompt2="Low quality"):
    save_lora_dir = "output"
    os.makedirs(save_lora_dir, exist_ok=True)
    
    # Load audio and compute duration
    waveform, sample_rate = torchaudio.load(audio_file1)
    duration = waveform.shape[1] / sample_rate
    duration = int(duration)
    
    # Perform morphing using the pipeline
    _ = pipeline(
        audio_file=audio_file1,
        audio_file2=audio_file2,
        audio_length_in_s=duration,
        time_pooling=2,
        freq_pooling=2,
        prompt_1=prompt1,
        prompt_2=prompt2,
        negative_prompt_1=negative_prompt1,
        negative_prompt_2=negative_prompt2,
        save_lora_dir=save_lora_dir,
        use_adain=True,
        use_reschedule=False,
        num_inference_steps=50,
        lamd=0.6,
        output_path=save_lora_dir,
        num_frames=5,
        fix_lora=None,
        use_lora=True,
        lora_steps=50,
        noisy_latent_with_lora=True,
        morphing_with_lora=True,
        use_morph_prompt=True,
        guidance_scale=7.5,
    )
    
    # Collect the output file paths
    output_paths = [os.path.join(save_lora_dir, file) for file in os.listdir(save_lora_dir) if file.endswith(".wav")]
    return output_paths

# Gradio interface function
def interface(audio1, audio2, prompt1, prompt2):
    output_paths = morph_audio(audio1, audio2, prompt1, prompt2)
    return output_paths

# Gradio Interface
demo = gr.Interface(
    fn=interface,
    inputs=[
        gr.Audio(label="Upload Audio File 1", type="filepath"),
        gr.Audio(label="Upload Audio File 2", type="filepath"),
        # gr.Slider(4, 6, step=1, label="Octave 1"),
        gr.Textbox(label="Prompt for Audio File 1"),
        gr.Textbox(label="Prompt for Audio File 2")
    ],
    outputs=[
        gr.Audio(label="Morphing audio 1"),
        gr.Audio(label="Morphing audio 2"),
        gr.Audio(label="Morphing audio 3"),
        gr.Audio(label="Morphing audio 4"),
        gr.Audio(label="Morphing audio 5"),
    ],
)

if __name__ == "__main__":
    demo.launch()