import gradio as gr from transformers import pipeline import tempfile, os from midi2audio import FluidSynth # --- Music Generation Logic (API-like Function) --- def generate_music_api(midi_data=None, chord_progression=None, tempo=120, temperature=0.95, nb_tokens=512, bar_range="0-4"): try: # Load the MusicLang Predict model (replace with actual loading code) ml = ... # Example: ml = pipeline("music-generation", model="your-musiclang-predict-model") # Handle different generation scenarios based on inputs if midi_data is not None and chord_progression.strip() != "": # Continue sequence with chord progression generated_score = ml.continue_sequence( midi_data, chord_progression=chord_progression, nb_tokens=int(nb_tokens), temperature=float(temperature), # ... other parameters ) elif midi_data is not None and chord_progression.strip() == "": # Generate using the uploaded MIDI file as a prompt generated_score = ml.predict( midi_data, # Use the uploaded MIDI as the prompt nb_tokens=int(nb_tokens), temperature=float(temperature), # ... other parameters ) else: # Generate with specific chord progression generated_score = ml.predict_chords( chord_progression, # ... other parameters ) # Save generated files to temporary locations temp_midi_file = tempfile.NamedTemporaryFile(suffix=".mid", delete=False) midi_path = temp_midi_file.name generated_score.to_midi(midi_path, tempo=tempo, time_signature=time_signature) # Assuming time_signature is defined temp_mp3_file = tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) mp3_path = temp_mp3_file.name # ... (convert MIDI to MP3 using FluidSynth and FFmpeg) # Read binary data from the temporary files with open(mp3_path, 'rb') as f_mp3: mp3_binary = f_mp3.read() with open(midi_path, 'rb') as f_midi: midi_binary = f_midi.read() # Remove temporary files os.remove(mp3_path) os.remove(midi_path) return { "mp3": mp3_binary, "midi": midi_binary, "chord_repr": chord_repr, # Assuming chord_repr is still needed "tempo_message": tempo_message # Assuming tempo_message is still needed } except Exception as e: return {"error": str(e)} # --- Gradio Interface --- def musiclang_gradio(midi_file, chord_progression, tempo, temperature, nb_tokens, bar_range): midi_data = None if midi_file: with open(midi_file.name, "rb") as f: midi_data = f.read() api_response = generate_music_api(midi_data=midi_data, chord_progression=chord_progression, tempo=tempo, temperature=temperature, nb_tokens=nb_tokens, bar_range=bar_range) if "error" in api_response: return None, None, api_response["error"] # Create temporary files for Gradio mp3_path = tempfile.NamedTemporaryFile(suffix=".mp3", delete=False).name midi_path = tempfile.NamedTemporaryFile(suffix=".mid", delete=False).name # Write binary data to temporary files with open(mp3_path, "wb") as f: f.write(api_response["mp3"]) with open(midi_path, "wb") as f: f.write(api_response["midi"]) return mp3_path, midi_path, None with gr.Blocks() as demo: # Introductory text gr.Markdown(""" # Controllable Symbolic Music Generation with MusicLang Predict [MusicLang Predict](https://github.com/musiclang/musiclang_predict) offers advanced controllability features and high-quality music generation by manipulating symbolic music. You can for example use it to continue your composition with a specific chord progression. """) with gr.Row(): with gr.Column(): with gr.Row(): midi_file = gr.File(label="Prompt MIDI File (Optional)", type="filepath", file_types=[".mid", ".midi"], elem_id='midi_file_input') with gr.Column(): bar_range = gr.Textbox(label="Bar Range of input file (eg: 0-4 for first four bars)", placeholder="0-4", value="0-4", elem_id='bar_range_input') nb_tokens = gr.Number(label="Nb Tokens", value=512, minimum=256, maximum=2048, step=256, elem_id='nb_tokens_input') temperature = gr.Slider( label="Temperature", value=0.95, visible=False, minimum=0.1, maximum=1.0, step=0.1, elem_id='temperature_input') tempo = gr.Slider(label="Tempo", value=120, minimum=60, maximum=240, step=1, elem_id='tempo_input') with gr.Row(): chord_progression = gr.Textbox( label="Chord Progression (Optional)", placeholder="Am CM Dm7/F E7 Asus4", lines=2, value="", elem_id='chord_progression_input') with gr.Row(): generate_btn = gr.Button("Generate", elem_id='generate_button') with gr.Column(): info_message = gr.Textbox(label="Info Message", elem_id='info_message_output') generated_music = gr.Audio(label="Preview generated Music", elem_id='generated_music_output') generated_midi = gr.File(label="Download MIDI", elem_id='generated_midi_output') generate_btn.click( fn=musiclang_gradio, inputs=[midi_file, chord_progression, tempo, temperature, nb_tokens, bar_range], outputs=[generated_music, generated_midi, info_message] ) with gr.Row(): with gr.Column(): gr.Markdown("## Examples") gr.Examples( examples=[["/home/user/app/bach_847.mid", "", 120, 0.95, 512, "0-4"], ["/home/user/app/bach_847.mid", "Cm C7/E Fm F#dim G7", 120, 0.95, 512, "0-4"], ["/home/user/app/boney_m_ma_baker.mid", "", 120, 0.95, 512, "0-4"], ["/home/user/app/eminem_slim_shady.mid", "Cm AbM BbM G7 Cm", 120, 0.95, 512, "0-4"], ["/home/user/app/mozart_alla_turca.mid", "", 120, 0.95, 512, "0-4"], ["/home/user/app/mozart_alla_turca.mid", "Am Em CM G7 E7 Am Am E7 Am", 120, 0.95, 512, "0-4"], ["/home/user/app/daft_punk_around_the_world.mid", "", 120, 0.95, 512, "0-4"], ], inputs=[midi_file, chord_progression, tempo, temperature, nb_tokens, bar_range], outputs=[generated_music, generated_midi, info_message], fn=musiclang_gradio, cache_examples=True, ) demo.launch()