import os import re import random from scipy.io.wavfile import write import gradio as gr from Applio import * from uvrmodel import * def roformer_separator(roformer_audio, roformer_model, roformer_output_format, roformer_overlap): files_list = [] files_list.clear() directory = "./outputs" random_id = str(random.randint(10000, 99999)) pattern = f"{random_id}" os.makedirs("outputs", exist_ok=True) write(f'{random_id}.wav', roformer_audio[0], roformer_audio[1]) full_roformer_model = roformer_models[roformer_model] prompt = f"audio-separator {random_id}.wav --model_filename {full_roformer_model} --output_dir=./outputs --output_format={roformer_output_format} --normalization=0.9 --mdxc_overlap={roformer_overlap}" os.system(prompt) for file in os.listdir(directory): if re.search(pattern, file): files_list.append(os.path.join(directory, file)) stem1_file = files_list[0] stem2_file = files_list[1] return stem1_file, stem2_file def mdxc_separator(mdx23c_audio, mdx23c_model, mdx23c_output_format, mdx23c_segment_size, mdx23c_overlap): files_list = [] files_list.clear() directory = "./outputs" random_id = str(random.randint(10000, 99999)) pattern = f"{random_id}" os.makedirs("outputs", exist_ok=True) write(f'{random_id}.wav', mdx23c_audio[0], mdx23c_audio[1]) prompt = f"audio-separator {random_id}.wav --model_filename {mdx23c_model} --output_dir=./outputs --output_format={mdx23c_output_format} --normalization=0.9 --mdxc_segment_size={mdx23c_segment_size} --mdxc_overlap={mdx23c_overlap}" os.system(prompt) for file in os.listdir(directory): if re.search(pattern, file): files_list.append(os.path.join(directory, file)) stem1_file = files_list[0] stem2_file = files_list[1] return stem1_file, stem2_file def mdxnet_separator(mdxnet_audio, mdxnet_model, mdxnet_output_format, mdxnet_segment_size, mdxnet_overlap, mdxnet_denoise): files_list = [] files_list.clear() directory = "./outputs" random_id = str(random.randint(10000, 99999)) pattern = f"{random_id}" os.makedirs("outputs", exist_ok=True) write(f'{random_id}.wav', mdxnet_audio[0], mdxnet_audio[1]) prompt = f"audio-separator {random_id}.wav --model_filename {mdxnet_model} --output_dir=./outputs --output_format={mdxnet_output_format} --normalization=0.9 --mdx_segment_size={mdxnet_segment_size} --mdx_overlap={mdxnet_overlap}" if mdxnet_denoise: prompt += " --mdx_enable_denoise" os.system(prompt) for file in os.listdir(directory): if re.search(pattern, file): files_list.append(os.path.join(directory, file)) stem1_file = files_list[0] stem2_file = files_list[1] return stem1_file, stem2_file def vrarch_separator(vrarch_audio, vrarch_model, vrarch_output_format, vrarch_window_size, vrarch_agression, vrarch_tta, vrarch_high_end_process): files_list = [] files_list.clear() directory = "./outputs" random_id = str(random.randint(10000, 99999)) pattern = f"{random_id}" os.makedirs("outputs", exist_ok=True) write(f'{random_id}.wav', vrarch_audio[0], vrarch_audio[1]) prompt = f"audio-separator {random_id}.wav --model_filename {vrarch_model} --output_dir=./outputs --output_format={vrarch_output_format} --normalization=0.9 --vr_window_size={vrarch_window_size} --vr_aggression={vrarch_agression}" if vrarch_tta: prompt += " --vr_enable_tta" if vrarch_high_end_process: prompt += " --vr_high_end_process" os.system(prompt) for file in os.listdir(directory): if re.search(pattern, file): files_list.append(os.path.join(directory, file)) stem1_file = files_list[0] stem2_file = files_list[1] return stem1_file, stem2_file def demucs_separator(demucs_audio, demucs_model, demucs_output_format, demucs_shifts, demucs_overlap): files_list = [] files_list.clear() directory = "./outputs" random_id = str(random.randint(10000, 99999)) pattern = f"{random_id}" os.makedirs("outputs", exist_ok=True) write(f'{random_id}.wav', demucs_audio[0], demucs_audio[1]) prompt = f"audio-separator {random_id}.wav --model_filename {demucs_model} --output_dir=./outputs --output_format={demucs_output_format} --normalization=0.9 --demucs_shifts={demucs_shifts} --demucs_overlap={demucs_overlap}" os.system(prompt) for file in os.listdir(directory): if re.search(pattern, file): files_list.append(os.path.join(directory, file)) stem1_file = files_list[0] stem2_file = files_list[1] stem3_file = files_list[2] stem4_file = files_list[3] return stem1_file, stem2_file, stem3_file, stem4_file with gr.Blocks(theme=applio, title="🎵 UVR5 UI 🎵") as app: gr.Markdown("

🎵 UVR5 UI 🎵

") gr.Markdown("If you liked this HF Space you can give me a ❤️") gr.Markdown("Try UVR5 UI with GPU using Colab [here](https://colab.research.google.com/github/Eddycrack864/UVR5-UI/blob/main/UVR_UI.ipynb)") with gr.Tabs(): with gr.TabItem("BS/Mel Roformer"): with gr.Row(): roformer_model = gr.Dropdown( label = "Select the Model", choices=list(roformer_models.keys()), interactive = True ) roformer_output_format = gr.Dropdown( label = "Select the Output Format", choices = output_format, interactive = True ) with gr.Row(): roformer_overlap = gr.Slider( minimum = 2, maximum = 4, step = 1, label = "Overlap", info = "Amount of overlap between prediction windows.", value = 4, interactive = True ) with gr.Row(): roformer_audio = gr.Audio( label = "Input Audio", type = "numpy", interactive = True ) with gr.Row(): roformer_button = gr.Button("Separate!", variant = "primary") with gr.Row(): roformer_stem1 = gr.Audio( show_download_button = True, interactive = False, label = "Stem 1", type = "filepath" ) roformer_stem2 = gr.Audio( show_download_button = True, interactive = False, label = "Stem 2", type = "filepath" ) roformer_button.click(roformer_separator, [roformer_audio, roformer_model, roformer_output_format, roformer_overlap], [roformer_stem1, roformer_stem2]) with gr.TabItem("MDX23C"): with gr.Row(): mdx23c_model = gr.Dropdown( label = "Select the Model", choices = mdx23c_models, interactive = True ) mdx23c_output_format = gr.Dropdown( label = "Select the Output Format", choices = output_format, interactive = True ) with gr.Row(): mdx23c_segment_size = gr.Slider( minimum = 32, maximum = 4000, step = 32, label = "Segment Size", info = "Larger consumes more resources, but may give better results.", value = 256, interactive = True ) mdx23c_overlap = gr.Slider( minimum = 2, maximum = 50, step = 1, label = "Overlap", info = "Amount of overlap between prediction windows.", value = 8, interactive = True ) with gr.Row(): mdx23c_audio = gr.Audio( label = "Input Audio", type = "numpy", interactive = True ) with gr.Row(): mdx23c_button = gr.Button("Separate!", variant = "primary") with gr.Row(): mdx23c_stem1 = gr.Audio( show_download_button = True, interactive = False, label = "Stem 1", type = "filepath" ) mdx23c_stem2 = gr.Audio( show_download_button = True, interactive = False, label = "Stem 2", type = "filepath" ) mdx23c_button.click(mdxc_separator, [mdx23c_audio, mdx23c_model, mdx23c_output_format, mdx23c_segment_size, mdx23c_overlap], [mdx23c_stem1, mdx23c_stem2]) with gr.TabItem("MDX-NET"): with gr.Row(): mdxnet_model = gr.Dropdown( label = "Select the Model", choices = mdxnet_models, interactive = True ) mdxnet_output_format = gr.Dropdown( label = "Select the Output Format", choices = output_format, interactive = True ) with gr.Row(): mdxnet_segment_size = gr.Slider( minimum = 32, maximum = 4000, step = 32, label = "Segment Size", info = "Larger consumes more resources, but may give better results.", value = 256, interactive = True ) mdxnet_overlap = gr.Dropdown( label = "Overlap", choices = mdxnet_overlap_values, value = mdxnet_overlap_values[0], interactive = True ) mdxnet_denoise = gr.Checkbox( label = "Denoise", info = "Enable denoising during separation.", value = True, interactive = True ) with gr.Row(): mdxnet_audio = gr.Audio( label = "Input Audio", type = "numpy", interactive = True ) with gr.Row(): mdxnet_button = gr.Button("Separate!", variant = "primary") with gr.Row(): mdxnet_stem1 = gr.Audio( show_download_button = True, interactive = False, label = "Stem 1", type = "filepath" ) mdxnet_stem2 = gr.Audio( show_download_button = True, interactive = False, label = "Stem 2", type = "filepath" ) mdxnet_button.click(mdxnet_separator, [mdxnet_audio, mdxnet_model, mdxnet_output_format, mdxnet_segment_size, mdxnet_overlap, mdxnet_denoise], [mdxnet_stem1, mdxnet_stem2]) with gr.TabItem("VR ARCH"): with gr.Row(): vrarch_model = gr.Dropdown( label = "Select the Model", choices = vrarch_models, interactive = True ) vrarch_output_format = gr.Dropdown( label = "Select the Output Format", choices = output_format, interactive = True ) with gr.Row(): vrarch_window_size = gr.Dropdown( label = "Window Size", choices = vrarch_window_size_values, value = vrarch_window_size_values[0], interactive = True ) vrarch_agression = gr.Slider( minimum = 1, maximum = 50, step = 1, label = "Agression", info = "Intensity of primary stem extraction.", value = 5, interactive = True ) vrarch_tta = gr.Checkbox( label = "TTA", info = "Enable Test-Time-Augmentation; slow but improves quality.", value = True, visible = True, interactive = True, ) vrarch_high_end_process = gr.Checkbox( label = "High End Process", info = "Mirror the missing frequency range of the output.", value = False, visible = True, interactive = True, ) with gr.Row(): vrarch_audio = gr.Audio( label = "Input Audio", type = "numpy", interactive = True ) with gr.Row(): vrarch_button = gr.Button("Separate!", variant = "primary") with gr.Row(): vrarch_stem1 = gr.Audio( show_download_button = True, interactive = False, type = "filepath", label = "Stem 1" ) vrarch_stem2 = gr.Audio( show_download_button = True, interactive = False, type = "filepath", label = "Stem 2" ) vrarch_button.click(vrarch_separator, [vrarch_audio, vrarch_model, vrarch_output_format, vrarch_window_size, vrarch_agression, vrarch_tta, vrarch_high_end_process], [vrarch_stem1, vrarch_stem2]) with gr.TabItem("Demucs"): with gr.Row(): demucs_model = gr.Dropdown( label = "Select the Model", choices = demucs_models, interactive = True ) demucs_output_format = gr.Dropdown( label = "Select the Output Format", choices = output_format, interactive = True ) with gr.Row(): demucs_shifts = gr.Slider( minimum = 1, maximum = 20, step = 1, label = "Shifts", info = "Number of predictions with random shifts, higher = slower but better quality.", value = 2, interactive = True ) demucs_overlap = gr.Dropdown( label = "Overlap", choices = demucs_overlap_values, value = demucs_overlap_values[0], interactive = True ) with gr.Row(): demucs_audio = gr.Audio( label = "Input Audio", type = "numpy", interactive = True ) with gr.Row(): demucs_button = gr.Button("Separate!", variant = "primary") with gr.Row(): demucs_stem1 = gr.Audio( show_download_button = True, interactive = False, type = "filepath", label = "Stem 1" ) demucs_stem2 = gr.Audio( show_download_button = True, interactive = False, type = "filepath", label = "Stem 2" ) with gr.Row(): demucs_stem3 = gr.Audio( show_download_button = True, interactive = False, type = "filepath", label = "Stem 3" ) demucs_stem4 = gr.Audio( show_download_button = True, interactive = False, type = "filepath", label = "Stem 4" ) demucs_button.click(demucs_separator, [demucs_audio, demucs_model, demucs_output_format, demucs_shifts, demucs_overlap], [demucs_stem1, demucs_stem2, demucs_stem3, demucs_stem4]) with gr.TabItem("Credits"): gr.Markdown( """ UVR5 UI created by **[Not Eddy (Spanish Mod)](http://discord.com/users/274566299349155851)** in **[AI HUB](https://discord.gg/aihub)** community. * python-audio-separator by [beveradb](https://github.com/beveradb). * Thanks to [Ilaria](https://github.com/TheStingerX) and [Mikus](https://github.com/cappuch) for the help with the code. * Improvements by [Blane187](https://github.com/Blane187). You can donate to the original UVR5 project here: [!["Buy Me A Coffee"](https://www.buymeacoffee.com/assets/img/custom_images/orange_img.png)](https://www.buymeacoffee.com/uvr5) """ ) app.queue() app.launch(show_api=False)