mosecanolimit / app /pages /Separate.py
fabiogra
fix: remove st_local_audio, convert orig audio to mp3 in inference script
78dd18a
raw
history blame
11.4 kB
import os
from pathlib import Path
from typing import List
from loguru import logger as log
import requests
import streamlit as st
from streamlit_option_menu import option_menu
from footer import footer
from header import header
from helpers import (
load_audio_segment,
load_list_of_songs,
plot_audio,
url_is_valid,
file_size_is_valid,
delete_old_files,
)
from service.demucs_runner import separator
from service.vocal_remover.runner import load_model, separate
from style import CSS_TABS
label_sources = {
"no_vocals.mp3": "๐ŸŽถ Instrumental",
"vocals.mp3": "๐ŸŽค Vocals",
"drums.mp3": "๐Ÿฅ Drums",
"bass.mp3": "๐ŸŽธ Bass",
"guitar.mp3": "๐ŸŽธ Guitar",
"piano.mp3": "๐ŸŽน Piano",
"other.mp3": "๐ŸŽถ Other",
}
separation_mode_to_model = {
"Vocals & Instrumental (Low Quality, Faster)": (
"vocal_remover",
["vocals.mp3", "no_vocals.mp3"],
),
"Vocals & Instrumental (High Quality, Slower)": ("htdemucs", ["vocals.mp3", "no_vocals.mp3"]),
"Vocals, Drums, Bass & Other (Slower)": (
"htdemucs",
["vocals.mp3", "drums.mp3", "bass.mp3", "other.mp3"],
),
"Vocal, Drums, Bass, Guitar, Piano & Other (Slowest)": (
"htdemucs_6s",
["vocals.mp3", "drums.mp3", "bass.mp3", "guitar.mp3", "piano.mp3", "other.mp3"],
),
}
extensions = ["mp3", "wav", "ogg", "flac"]
out_path = Path("/tmp")
in_path = Path("/tmp")
@st.cache_data(show_spinner=False)
def get_sources(path, file_sources):
sources = {}
for file in file_sources:
fullpath = path / file
if fullpath.exists():
sources[file] = fullpath
return sources
def reset_execution():
st.session_state.executed = False
def show_results(model_name: str, dir_name_output: str, file_sources: List):
sources = get_sources(out_path / Path(model_name) / dir_name_output, file_sources)
tab_sources = st.tabs([f"**{label_sources.get(k)}**" for k in sources.keys()])
for i, (file, pathname) in enumerate(sources.items()):
with tab_sources[i]:
cols = st.columns(2)
with cols[0]:
auseg = load_audio_segment(pathname, "mp3")
st.image(
plot_audio(
auseg,
32767,
file=file,
model_name=model_name,
dir_name_output=dir_name_output,
),
use_column_width="always",
)
with cols[1]:
st.audio(str(pathname))
log.info(f"Displaying results for {dir_name_output} - {model_name}")
def body():
filename = None
name_song = None
st.markdown(
"<h4><center>Extract Vocals & Instrumental from any song</center></h4>",
unsafe_allow_html=True,
)
st.markdown(CSS_TABS, unsafe_allow_html=True)
cols = st.columns([1, 4, 1, 3, 1])
with cols[1]:
with st.columns([1, 8, 1])[1]:
option = option_menu(
menu_title=None,
options=["Examples", "Upload File", "From URL"],
icons=["cloud-upload-fill", "link-45deg", "music-note-list"],
orientation="horizontal",
styles={
"container": {
"width": "100%",
"height": "3.5rem",
"margin": "0px",
"padding": "0px",
},
"icon": {"font-size": "1rem"},
"nav-link": {
"display": "flex",
"height": "3rem",
"justify-content": "center",
"align-items": "center",
"text-align": "center",
"flex-direction": "column",
"font-size": "1rem",
"padding-left": "0px",
"padding-right": "0px",
},
},
key="option_separate",
)
if option == "Examples":
samples_song = load_list_of_songs(path="separate_songs.json")
if samples_song is not None:
name_song = st.selectbox(
label="Select a sample song and listen to sources separated",
options=list(samples_song.keys()) + [""],
format_func=lambda x: x.replace("_", " "),
index=len(samples_song),
key="select_example",
)
full_path = f"{in_path}/{name_song}"
if name_song != "" and Path(full_path).exists():
st.audio(full_path)
else:
name_song = None
elif option == "Upload File":
uploaded_file = st.file_uploader(
"Choose a file",
type=extensions,
key="file",
help="Supported formats: mp3, wav, ogg, flac.",
)
if uploaded_file is not None:
with st.spinner("Loading audio..."):
with open(in_path / uploaded_file.name, "wb") as f:
f.write(uploaded_file.getbuffer())
filename = uploaded_file.name
st.audio(uploaded_file)
elif option == "From URL":
url = st.text_input(
"Paste the URL of the audio file",
key="url_input",
help="Supported formats: mp3, wav, ogg, flac.",
)
if url != "" and url_is_valid(url):
with st.spinner("Downloading audio..."):
filename = url.split("/")[-1]
response = requests.get(url, stream=True)
if response.status_code == 200 and file_size_is_valid(
response.headers.get("Content-Length")
):
file_size = 0
with open(in_path / filename, "wb") as audio_file:
for chunk in response.iter_content(chunk_size=1024):
if chunk:
audio_file.write(chunk)
file_size += len(chunk)
if not file_size_is_valid(file_size):
audio_file.close()
os.remove(in_path / filename)
filename = None
return
st.audio(f"{in_path}/{filename}")
else:
st.error(
"Failed to download audio file. Try to download it manually and upload it."
)
filename = None
with cols[3]:
separation_mode = st.selectbox(
"Choose the separation mode",
[
"Vocals & Instrumental (Low Quality, Faster)",
"Vocals & Instrumental (High Quality, Slower)",
"Vocals, Drums, Bass & Other (Slower)",
"Vocal, Drums, Bass, Guitar, Piano & Other (Slowest)",
],
on_change=reset_execution(),
key="separation_mode",
)
if separation_mode == "Vocals & Instrumental (Low Quality, Faster)":
max_duration = 30
else:
max_duration = 15
model_name, file_sources = separation_mode_to_model[separation_mode]
if filename is not None:
song = load_audio_segment(in_path / filename, filename.split(".")[-1])
n_secs = round(len(song) / 1000)
if os.environ.get("ENV_LIMITATION", False):
with cols[3]:
start_time = st.number_input(
"Choose the start time",
min_value=0,
max_value=n_secs,
step=1,
value=0,
help=f"Maximum duration is {max_duration} seconds for this separation mode.\nDuplicate this space to [remove any limit](https://github.com/fabiogra/moseca#are-there-any-limitations).",
format="%d",
)
st.session_state.start_time = start_time
end_time = min(start_time + max_duration, n_secs)
song = song[start_time * 1000 : end_time * 1000]
st.info(
f"Audio source will be processed from {start_time} to {end_time} seconds.\nDuplicate this space to [remove any limit](https://github.com/fabiogra/moseca#are-there-any-limitations).",
icon="โฑ",
)
else:
start_time = 0
end_time = n_secs
with st.columns([2, 1, 2])[1]:
execute = st.button(
"Separate Music Sources ๐ŸŽถ", type="primary", use_container_width=True
)
if execute or st.session_state.executed:
if execute:
st.session_state.executed = False
if not st.session_state.executed:
log.info(f"{option} - Separating {filename} with {separation_mode}...")
song.export(in_path / filename, format=filename.split(".")[-1])
with st.columns([1, 1, 1])[1]:
with st.spinner("Separating source audio, it will take a while..."):
if model_name == "vocal_remover":
model, device = load_model(pretrained_model="baseline.pth")
separate(
input=in_path / filename,
model=model,
device=device,
output_dir=out_path,
)
else:
stem = None
if separation_mode == "Vocals & Instrumental (High Quality, Slower)":
stem = "vocals"
separator(
tracks=[in_path / filename],
out=out_path,
model=model_name,
shifts=1,
overlap=0.5,
stem=stem,
int24=False,
float32=False,
clip_mode="rescale",
mp3=True,
mp3_bitrate=320,
verbose=True,
start_time=start_time,
end_time=end_time,
)
dir_name_output = ".".join(filename.split(".")[:-1])
filename = None
st.session_state.executed = True
show_results(model_name, dir_name_output, file_sources)
elif name_song is not None and option == "Examples":
show_results(model_name, name_song, file_sources)
if __name__ == "__main__":
header()
body()
footer()
delete_old_files("/tmp", 60 * 30)