import os import json import shutil import argparse import warnings import gradio as gr from generate import generate_music, get_args from utils import WEIGHTS_DIR, TEMP_DIR, LANG EN2ZH = { "Cite": "引用", "Submit": "提交", "Feedback: the emotion you believe the generated result should belong to": "反馈：你所认为的生成结果该所属的情感", "Status": "状态栏", "Staff": "五线谱", "ABC notation": "ABC 记谱", "Download MXL": "下载 MXL", "Download MusicXML": "下载 MusicXML", "Download PDF score": "下载 PDF 乐谱", "Download MIDI": "下载 MIDI", "Audio": "音频", "Download template": "下载模板", "Save template": "保存模板", "The emotion to which the current template belongs": "当前模板所属情感", "Generate": "生成", "Generate chords coming soon": "生成和声控制暂不可用", "Volume in dB": "dB 音量调节", "±12 octave": "±12 八度上下移", "BPM tempo": "BPM 速度", "Minor": "小调", "Major": "大调", "Mode": "大小调", "Pitch SD": "音高标准差", "Low": "低", "High": "高", "By feature control": "通过特征控制生成", "By template": "通过模板生成", "Arousal: reflects the calmness-intensity of the emotion": "唤醒度 反映情绪的 平静-激烈 程度", "Valence: reflects negative-positive levels of emotion": "愉悦度 反映情绪的 消极-积极 程度", "Video demo": "视频教程", "Dataset": "数据集", "Status": "状态栏", } def _L(en_txt: str): return en_txt if LANG else f"{en_txt} ({EN2ZH[en_txt]})" def infer_by_template(dataset: str, v: str, a: str, add_chord: bool): status = "Success" audio = midi = pdf = xml = mxl = tunes = jpg = None emotion = "Q1" if v == _L("Low") and a == _L("High"): emotion = "Q2" elif v == _L("Low") and a == _L("Low"): emotion = "Q3" elif v == _L("High") and a == _L("Low"): emotion = "Q4" try: parser = argparse.ArgumentParser() args = get_args(parser) args.template = True audio, midi, pdf, xml, mxl, tunes, jpg = generate_music( args, emo=emotion, weights=f"{WEIGHTS_DIR}/{dataset.lower()}/weights.pth", ) except Exception as e: status = f"{e}" return status, audio, midi, pdf, xml, mxl, tunes, jpg def infer_by_features( dataset: str, pitch_std: str, mode: str, tempo: int, octave: int, rms: int, add_chord: bool, ): status = "Success" audio = midi = pdf = xml = mxl = tunes = jpg = None emotion = "Q1" if mode == _L("Minor") and pitch_std == _L("High"): emotion = "Q2" elif mode == _L("Minor") and pitch_std == _L("Low"): emotion = "Q3" elif mode == _L("Major") and pitch_std == _L("Low"): emotion = "Q4" try: parser = argparse.ArgumentParser() args = get_args(parser) args.template = False audio, midi, pdf, xml, mxl, tunes, jpg = generate_music( args, emo=emotion, weights=f"{WEIGHTS_DIR}/{dataset.lower()}/weights.pth", fix_tempo=tempo, fix_pitch=octave, fix_volume=rms, ) except Exception as e: status = f"{e}" return status, audio, midi, pdf, xml, mxl, tunes, jpg def feedback( fixed_emo: str, source_dir=f"./{TEMP_DIR}/output", target_dir=f"./{TEMP_DIR}/feedback", ): try: if not fixed_emo: raise ValueError("Please select feedback before submitting! ") os.makedirs(target_dir, exist_ok=True) for root, _, files in os.walk(source_dir): for file in files: if file.endswith(".mxl"): prompt_emo = file.split("]")[0][1:] if prompt_emo != fixed_emo: file_path = os.path.join(root, file) target_path = os.path.join( target_dir, file.replace(".mxl", f"_{fixed_emo}.mxl") ) shutil.copy(file_path, target_path) return f"Copied {file_path} to {target_path}" else: return "Thanks for your feedback!" return "No .mxl files found in the source directory." except Exception as e: return f"{e}" def save_template(label: str, pitch_std: str, mode: str, tempo: int, octave: int, rms): status = "Success" template = None try: if ( label and pitch_std and mode and tempo != None and octave != None and rms != None ): json_str = json.dumps( { "label": label, "pitch_std": pitch_std == _L("High"), "mode": mode == _L("Major"), "tempo": tempo, "octave": octave, "volume": rms, } ) with open( f"./{TEMP_DIR}/feedback/templates.jsonl", "a", encoding="utf-8", ) as file: file.write(json_str + "\n") template = f"./{TEMP_DIR}/feedback/templates.jsonl" else: raise ValueError("Please check features") except Exception as e: status = f"{e}" return status, template if __name__ == "__main__": warnings.filterwarnings("ignore") with gr.Blocks() as demo: if LANG: gr.Markdown( "## The current CPU-based version on HuggingFace has slow inference, you can access the GPU-based mirror on [ModelScope](https://www.modelscope.cn/studios/monetjoe/EMelodyGen)" ) with gr.Row(): with gr.Column(): gr.Video( "./demo.mp4" if LANG else "./src/tutorial.mp4", label=_L("Video demo"), show_download_button=False, show_share_button=False, ) dataset_option = gr.Dropdown( ["VGMIDI", "EMOPIA", "Rough4Q"], label=_L("Dataset"), value="Rough4Q", ) with gr.Tab(_L("By template")): gr.Image( "https://www.modelscope.cn/studio/monetjoe/EMelodyGen/resolve/master/src/4q.jpg", show_label=False, show_download_button=False, show_fullscreen_button=False, show_share_button=False, ) valence_radio = gr.Radio( [_L("Low"), _L("High")], label=_L( "Valence: reflects negative-positive levels of emotion" ), value=_L("High"), ) arousal_radio = gr.Radio( [_L("Low"), _L("High")], label=_L( "Arousal: reflects the calmness-intensity of the emotion" ), value=_L("High"), ) chord_check = gr.Checkbox( label=_L("Generate chords coming soon"), value=False, ) gen_btn_1 = gr.Button(_L("Generate")) with gr.Tab(_L("By feature control")): std_option = gr.Radio( [_L("Low"), _L("High")], label=_L("Pitch SD"), value=_L("High") ) mode_option = gr.Radio( [_L("Minor"), _L("Major")], label=_L("Mode"), value=_L("Major") ) tempo_option = gr.Slider( minimum=40, maximum=228, step=1, value=120, label=_L("BPM tempo"), ) octave_option = gr.Slider( minimum=-24, maximum=24, step=12, value=0, label=_L("±12 octave"), ) volume_option = gr.Slider( minimum=-5, maximum=10, step=5, value=0, label=_L("Volume in dB"), ) chord_check_2 = gr.Checkbox( label=_L("Generate chords coming soon"), value=False, ) gen_btn_2 = gr.Button(_L("Generate")) template_radio = gr.Radio( ["Q1", "Q2", "Q3", "Q4"], label=_L("The emotion to which the current template belongs"), ) save_btn = gr.Button(_L("Save template")) dld_template = gr.File(label=_L("Download template")) with gr.Column(): wav_audio = gr.Audio(label=_L("Audio"), type="filepath") midi_file = gr.File(label=_L("Download MIDI")) pdf_file = gr.File(label=_L("Download PDF score")) xml_file = gr.File(label=_L("Download MusicXML")) mxl_file = gr.File(label=_L("Download MXL")) abc_textbox = gr.Textbox( label=_L("ABC notation"), show_copy_button=True ) staff_img = gr.Image(label=_L("Staff"), type="filepath") with gr.Column(): status_bar = gr.Textbox(label=_L("Status"), show_copy_button=True) fdb_radio = gr.Radio( ["Q1", "Q2", "Q3", "Q4"], label=_L( "Feedback: the emotion you believe the generated result should belong to" ), ) fdb_btn = gr.Button(_L("Submit")) gr.Markdown( f"""## {_L("Cite")} ```bibtex @inproceedings{{Zhou2025EMelodyGen, title = {{EMelodyGen: Emotion-Conditioned Melody Generation in ABC Notation with the Musical Feature Template}}, author = {{Monan Zhou and Xiaobing Li and Feng Yu and Wei Li}}, month = {{Mar}}, year = {{2025}}, publisher = {{GitHub}}, version = {{0.1}}, url = {{https://github.com/monetjoe/EMelodyGen}} }} ```""" ) # actions gen_btn_1.click( fn=infer_by_template, inputs=[dataset_option, valence_radio, arousal_radio, chord_check], outputs=[ status_bar, wav_audio, midi_file, pdf_file, xml_file, mxl_file, abc_textbox, staff_img, ], ) gen_btn_2.click( fn=infer_by_features, inputs=[ dataset_option, std_option, mode_option, tempo_option, octave_option, volume_option, chord_check, ], outputs=[ status_bar, wav_audio, midi_file, pdf_file, xml_file, mxl_file, abc_textbox, staff_img, ], ) save_btn.click( fn=save_template, inputs=[ template_radio, std_option, mode_option, tempo_option, octave_option, volume_option, ], outputs=[status_bar, dld_template], ) fdb_btn.click(fn=feedback, inputs=fdb_radio, outputs=status_bar) demo.launch()