en:
  # Translation by Eddycrack864
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Select the model
  Select the output format : Select the output format
  Overlap : Overlap
  Amount of overlap between prediction windows : Amount of overlap between prediction windows
  Segment size : Segment size
  Larger consumes more resources, but may give better results : Larger consumes more resources, but may give better results
  Input audio : Input audio
  Separation by link : Separation by link
  Link : Link
  Paste the link here : Paste the link here
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Download!
  Batch separation : Batch separation
  Input path : Input path
  Place the input path here : Place the input path here
  Output path : Output path
  Place the output path here : Place the output path here
  Separate! : Separate!
  Output information : Output information
  Stem 1 : Stem 1
  Stem 2 : Stem 2
  Denoise : Denoise
  Enable denoising during separation : Enable denoising during separation
  Window size : Window size
  Agression : Agression
  Intensity of primary stem extraction : Intensity of primary stem extraction
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Enable Test-Time-Augmentation; slow but improves quality
  High end process : High end process
  Mirror the missing frequency range of the output : Mirror the missing frequency range of the output
  Shifts : Shifts
  Number of predictions with random shifts, higher = slower but better quality : Number of predictions with random shifts, higher = slower but better quality
  Stem 3 : Stem 3
  Stem 4 : Stem 4
  Themes : Themes
  Theme : Theme
  Select the theme you want to use. (Requires restarting the App) : Select the theme you want to use. (Requires restarting the App)
  Credits : Credits
  Language : Language
  Advanced settings : Advanced settings
  Override model default segment size instead of using the model default value : Override model default segment size instead of using the model default value
  Override segment size : Override segment size
  Batch size : Batch size
  Larger consumes more RAM but may process slightly faster : Larger consumes more RAM but may process slightly faster
  Normalization threshold : Normalization threshold
  The threshold for audio normalization : The threshold for audio normalization
  Amplification threshold : Amplification threshold
  The threshold for audio amplification : The threshold for audio amplification
  Hop length : Hop length
  Usually called stride in neural networks; only change if you know what you're doing : Usually called stride in neural networks; only change if you know what you're doing
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality
  Identify leftover artifacts within vocal output; may improve separation for some songs : Identify leftover artifacts within vocal output; may improve separation for some songs
  Post process : Post process
  Post process threshold : Post process threshold
  Threshold for post-processing : Threshold for post-processing
  Size of segments into which the audio is split. Higher = slower but better quality : Size of segments into which the audio is split. Higher = slower but better quality
  Enable segment-wise processing : Enable segment-wise processing
  Segment-wise processing : Segment-wise processing
  Stem 5 : Stem 5
  Stem 6 : Stem 6

es:
  # Translation by Eddycrack864
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : Si te gusta UVR5 UI puedes darle una estrella a mi repo en [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Prueba UVR5 UI en Hugging Face con una A100 [aquí](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Selecciona el modelo
  Select the output format : Selecciona el formato de salida
  Overlap : Superposición
  Amount of overlap between prediction windows : Cantidad de superposición entre ventanas de predicción
  Segment size : Tamaño del segmento
  Larger consumes more resources, but may give better results : Un tamaño más grande consume más recursos, pero puede dar mejores resultados
  Input audio : Audio de entrada
  Separation by link : Separación por link
  Link : Link
  Paste the link here : Pega el link aquí
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : Puedes pegar el enlace al video/audio desde muchos sitios, revisa la lista completa [aquí](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Descargar!
  Batch separation : Separación por lotes
  Input path : Ruta de entrada
  Place the input path here : Coloca la ruta de entrada aquí
  Output path : Ruta de salida
  Place the output path here : Coloca la ruta de salida aquí
  Separate! : Separar!
  Output information : Información de la salida
  Stem 1 : Pista 1
  Stem 2 : Pista 2
  Denoise : Eliminación de ruido
  Enable denoising during separation : Habilitar la eliminación de ruido durante la separación
  Window size : Tamaño de la ventana
  Agression : Agresión
  Intensity of primary stem extraction : Intensidad de extracción de la pista primaria
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Habilitar Aumento del Tiempo de Prueba; lento pero mejora la calidad
  High end process : Procesamiento de alto rendimiento
  Mirror the missing frequency range of the output : Reflejar el rango de frecuencia faltante de la salida
  Shifts : Desplazamientos temporales
  Number of predictions with random shifts, higher = slower but better quality : Número de predicciones con desplazamientos temporales, mayor = más lento pero mejor calidad
  Stem 3 : Pista 3
  Stem 4 : Pista 4
  Themes : Temas
  Theme : Tema
  Select the theme you want to use. (Requires restarting the App) : Selecciona el tema que deseas utilizar. (Requiere reiniciar la aplicación)
  Credits : Créditos
  Language : Idioma
  Advanced settings : Configuración avanzada
  Override model default segment size instead of using the model default value : Anular el tamaño del segmento predeterminado del modelo en lugar de usar el valor predeterminado del modelo
  Override segment size : Anular tamaño del segmento
  Batch size : Tamaño del lote
  Larger consumes more RAM but may process slightly faster : Más grande consume más RAM pero puede procesar un poco más rápido
  Normalization threshold : Umbral de normalización
  The threshold for audio normalization : El umbral para la normalización del audio
  Amplification threshold : Umbral de amplificación
  The threshold for audio amplification : El umbral para la amplificación de audio
  Hop length : Longitud del salto
  Usually called stride in neural networks; only change if you know what you're doing : Generalmente llamado paso en redes neuronales; solo cambialo si sabes lo que estás haciendo
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : Equilibra la calidad y la velocidad. 1024 = más rápido pero de menor calidad, 320 = más lento pero de mejor calidad
  Identify leftover artifacts within vocal output; may improve separation for some songs : Identifica artefactos sobrantes en la salida vocal; puede mejorar la separación de algunas canciones
  Post process : Posproceso
  Post process threshold : Umbral de posproceso
  Threshold for post-processing : Umbral para el posprocesamiento
  Size of segments into which the audio is split. Higher = slower but better quality : Tamaño de los segmentos en los que se divide el audio. Más alto = más lento pero de mejor calidad
  Enable segment-wise processing : Habilitar el procesamiento por segmentos
  Segment-wise processing : Procesamiento por segmentos
  Stem 5 : Pista 5
  Stem 6 : Pista 6

it:
  # Thanks to Nick088 for the Italian translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : Se ti piace UVR5 UI puoi aggiungere una stella al mio repository su [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Prova UVR5 UI su Hugging Face con A100 [qui](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Seleziona il modello
  Select the output format : Seleziona il formato di output
  Overlap : Sovrapposizione
  Amount of overlap between prediction windows : Quantità di sovrapposizione tra le finestre di predizione
  Segment size : Dimensione del segmento
  Larger consumes more resources, but may give better results : Dimensioni maggiori consumano più risorse, ma potrebbero dare risultati migliori
  Input audio : Audio di input
  Separation by link : Separazione tramite link
  Link : Link
  Paste the link here : Incolla il link qui
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : Puoi incollare il link al video/audio da molti siti, controlla la lista completa [qui](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Scarica!
  Batch separation : Separazione in batch
  Input path : Percorso di input
  Place the input path here : Inserisci il percorso di input qui
  Output path : Percorso di output
  Place the output path here : Inserisci il percorso di output qui
  Separate! : Separa!
  Output information : Informazioni di output
  Stem 1 : Traccia 1
  Stem 2 : Traccia 2
  Denoise : Riduzione del rumore
  Enable denoising during separation : Abilita la riduzione del rumore durante la separazione
  Window size : Dimensione della finestra
  Agression : Aggressività
  Intensity of primary stem extraction : Intensità dell'estrazione della traccia primaria
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Abilita l'Aumento del Tempo di Prova; lento ma migliora la qualità
  High end process : Elaborazione ad alte prestazion
  Mirror the missing frequency range of the output : Rifletti l'intervallo di frequenze mancante dell'output
  Shifts : Spostamenti
  Number of predictions with random shifts, higher = slower but better quality : Numero di predizioni con spostamenti casuali, maggiore = più lento ma qualità migliore
  Stem 3 : Traccia 3
  Stem 4 : Traccia 4
  Themes : Temi
  Theme : Tema
  Select the theme you want to use. (Requires restarting the App) : Seleziona il tema che desideri utilizzare. (Richiede il riavvio dell'app)
  Credits : Crediti
  Language : Lingua
  Advanced settings : Impostazioni avanzate
  Override model default segment size instead of using the model default value : Sovrascrivi la dimensione di segmento predefinita del modello invece di utilizzare il valore predefinito del modello
  Override segment size : Ignora dimensione segmento
  Batch size : Dimensione del batch
  Larger consumes more RAM but may process slightly faster : Più grande consuma più RAM ma potrebbe elaborare leggermente più velocemente
  Normalization threshold : Soglia di normalizzazione
  The threshold for audio normalization : La soglia per la normalizzazione dell'audio
  Amplification threshold : Soglia di amplificazione
  The threshold for audio amplification : La soglia per l'amplificazione dell'audio
  Hop length : Lunghezza del salto
  Usually called stride in neural networks; only change if you know what you're doing : Solitamente chiamato passo nelle reti neurali; cambialo solo se sai cosa stai facendo
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : Bilancia la qualità e la velocità. 1024 = veloce ma inferiore, 320 = più lento ma migliore qualità
  Identify leftover artifacts within vocal output; may improve separation for some songs : Identifica gli artefatti residui nell'output vocale; potrebbe migliorare la separazione per alcune canzoni
  Post process : Post-elaborazione
  Post process threshold : Soglia di post-elaborazione
  Threshold for post-processing : Soglia per la post-elaborazione
  Size of segments into which the audio is split. Higher = slower but better quality : Dimensione dei segmenti in cui l'audio è diviso. Più alto = più lento ma migliore qualità
  Enable segment-wise processing : Abilita l'elaborazione per segmenti
  Segment-wise processing : Elaborazione per segmenti
  Stem 5 : Traccia 5
  Stem 6 : Traccia 6

pt:
  # Thanks to MrM0dZ for the Portuguese (Brazil) translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : Se você gosta do UVR5 UI, você pode favoritar meu repo em [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Tente UVR5 UI no Hugging Face com A100 [aqui](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Selecione o modelo
  Select the output format : Selecione o formato de saída.
  Overlap : Sobreposição
  Amount of overlap between prediction windows : Quantidade de sobreposição entre janelas de previsão
  Segment size : Tamanho de segmento
  Larger consumes more resources, but may give better results : Maior consume mais recursos, mas retorna melhores resultados
  Input audio : Áudio de entrada
  Separation by link : Separação por link
  Link : Link
  Paste the link here : Cole o link aqui
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : Você pode colar o link de um vídeo/áudio de vários sites, confira a lista [aqui](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Download!
  Batch separation : Separação de lote
  Input path : Caminho de entrada
  Place the input path here : Coloque o caminho de entrada aqui
  Output path : Caminho de saída
  Place the output path here : Coloque o caminho de saída aqui
  Separate! : Separar!
  Output information : Informação de saída
  Stem 1 : Stem 1
  Stem 2 : Stem 2
  Denoise : Reduçao de ruído
  Enable denoising during separation : Ativar redução de ruído durante separação
  Window size : Tamanho da janela
  Agression : Agressividade
  Intensity of primary stem extraction : Intensidade da extração de stem primaria
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Aumentar tempo de teste; lento mas melhora qualidade
  High end process : Processo de alta qualidade
  Mirror the missing frequency range of the output : Espelhar a frequência faltante de saida.
  Shifts : Turnos
  Number of predictions with random shifts, higher = slower but better quality : Numero de previsões com turnos aleatorios, maior = mais lento porem mais qualidade
  Stem 3 : Stem 3
  Stem 4 : Stem 4
  Themes : Temas
  Theme : Tema
  Select the theme you want to use. (Requires restarting the App) : Selecione o tema que deseja utilizar. (Requer reiniciar o App)
  Credits : Créditos
  Language : Idioma
  Advanced settings : Opções Avançadas
  Override model default segment size instead of using the model default value :  Substituir tamanho de segmento padrão ao invés de usar valor padrão do modelo
  Override segment size : Substituir tamanho de segmento
  Batch size : Tamanho do lote
  Larger consumes more RAM but may process slightly faster : Maior consome mais RAM, mas processa mais rapido.
  Normalization threshold : Limite de normalização
  The threshold for audio normalization : Limite de normalização para áudio
  Amplification threshold : Limite para amplicação
  The threshold for audio amplification : Limite para amplicação para áudio
  Hop length : Tamanho do pulo
  Usually called stride in neural networks; only change if you know what you're doing :  Normalmente chamado stride em redes neurais; Somente altere se souber o que está fazendo.
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality :  Balancear qualidade e velocidade. 1024 = Rápido porem lento, 320 = Lento porem melhor qualidade
  Identify leftover artifacts within vocal output; may improve separation for some songs : Identificar artefatos restantes na saída de vocal; Pode melhorar isolamento para algumas músicas.
  Post process : Pós-processamento
  Post process threshold : Limite de pós-processamento
  Threshold for post-processing : Limite para pós-processamento
  Size of segments into which the audio is split. Higher = slower but better quality : Tamanho de segmentos para cortar o áudio. Maior = Lento porem melhor qualidade
  Enable segment-wise processing : Ativar Processamento por segmento
  Segment-wise processing : Processamento por segmento
  Stem 5 : Stem 5
  Stem 6 : Stem 6

ms:
  # Thanks to Ryouko-Yamanda65777 for the Malay translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : Jika anda suka UI UVR5, anda boleh bintang repositori saya di [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Cuba UI UVR5 pada Memeluk Wajah dengan A100 [di sini](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Pilih model
  Select the output format : Pilih format output
  Overlap : Pertindihan
  Amount of overlap between prediction windows : Jumlah pertindihan antara tetingkap ramalan
  Segment size : Saiz segmen
  Larger consumes more resources, but may give better results : Saiz besar menggunakan lebih banyak sumber, tetapi mungkin memberikan hasil yang lebih baik
  Input audio : Input audio
  Separation by link : Pemisahan mengikut pautan
  Link : Pautan
  Paste the link here : Tampal pautan di sini
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : Anda boleh tampal pautan ke video/audio dari banyak laman, semak senarai lengkap [di sini](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Muat turun!
  Batch separation : Pemisahan kelompok
  Input path : Laluan input
  Place the input path here : Letakkan laluan input di sini
  Output path : Laluan output
  Place the output path here : Letakkan laluan output di sini
  Separate! : Pisahkan!
  Output information : Maklumat output
  Stem 1 : Lapis 1
  Stem 2 : Lapis 2
  Denoise : Nyahbunyi
  Enable denoising during separation : Aktifkan nyahbunyi semasa pemisahan
  Window size: Saiz tetingkap
  Agression: Kekasaran
  Intensity of primary stem extraction : Intensiti pengekstrakan lapis utama
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Aktifkan Augmentasi Ujian Masa; perlahan tetapi meningkatkan kualiti
  High end process : Proses hujung tinggi
  Mirror the missing frequency range of the output : Cerminkan julat frekuensi yang hilang pada output
  Shifts : Peralihan
  Number of predictions with random shifts, higher = slower but better quality : Bilangan ramalan dengan peralihan rawak, lebih tinggi = lebih perlahan tetapi kualiti lebih baik
  Stem 3 : Lapis 3
  Stem 4 : Lapis 4
  Themes : Tema
  Theme : Tema
  Select the theme you want to use. (Requires restarting the App) : Pilih tema yang anda mahu gunakan. (Perlu mulakan semula Aplikasi)
  Credits : Kredit
  Language : Bahasa
  Advanced settings: Tetapan lanjutan
  Override model default segment size instead of using the model default value: Gantikan saiz segmen lalai model daripada menggunakan nilai lalai model
  Override segment size: Gantikan saiz segmen
  Batch size: Saiz Batch
  Larger consumes more RAM but may process slightly faster: Saiz lebih besar menggunakan lebih banyak RAM tetapi mungkin memproses dengan lebih cepat
  Normalization threshold: Ambang normalisasi
  The threshold for audio normalization: Ambang untuk normalisasi audio
  Amplification threshold: Ambang amplifikasi
  The threshold for audio amplification: Ambang untuk amplifikasi audio
  Hop length: Panjang Hop
  Usually called stride in neural networks; only change if you know what you're doing: Biasanya dipanggil langkah dalam rangkaian neural; hanya ubah jika anda tahu apa yang anda lakukan
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality: Imbangkan kualiti dan kelajuan. 1024 = cepat tetapi lebih rendah, 320 = lebih perlahan tetapi kualiti lebih baik
  Identify leftover artifacts within vocal output; may improve separation for some songs: Kenalpasti artifak sisa dalam keluaran vokal; boleh meningkatkan pemisahan untuk sesetengah lagu
  Post process: Proses pasca
  Post process threshold: Ambang proses pasca
  Threshold for post-processing: Ambang untuk pemprosesan pasca
  Size of segments into which the audio is split. Higher = slower but better quality: Saiz segmen di mana audio dipecahkan. Lebih besar = lebih perlahan tetapi kualiti lebih baik
  Enable segment-wise processing: Aktifkan pemprosesan mengikut segmen
  Segment-wise processing: Pemprosesan mengikut segmen
  Stem 5: Lapis 5
  Stem 6: Lapis 6

id:
  # Thanks to Ryouko-Yamanda65777 for the Indonesian translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : Jika Anda suka UI UVR5, Anda dapat memberikan bintang pada repo saya di [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Coba UVR5 UI di Hugging Face dengan A100 [di sini](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Pilih model
  Select the output format : Pilih format output
  Overlap : Tumpang tindih
  Amount of overlap between prediction windows : Jumlah tumpang tindih antara jendela prediksi
  Segment size : Ukuran segmen
  Larger consumes more resources, but may give better results : Lebih besar menggunakan lebih banyak sumber daya, tetapi dapat memberikan hasil yang lebih baik
  Input audio : Input audio
  Separation by link : Pemisahan berdasarkan tautan
  Link : Tautan
  Paste the link here : Tempel tautan di sini
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : Anda dapat menempelkan tautan ke video/audio dari banyak situs, lihat daftar lengkap [di sini](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Unduh!
  Batch separation : Pemisahan batch
  Input path : Jalur input
  Place the input path here : Letakkan jalur input di sini
  Output path : Jalur output
  Place the output path here : Tempatkan jalur output di sini
  Separate! : Pisahkan!
  Output information : Informasi output
  Stem 1 : Stem 1
  Stem 2 : Stem 2
  Denoise : Mengurangi kebisingan
  Enable denoising during separation : Aktifkan pengurangan kebisingan selama pemisahan
  Window size : Ukuran jendela
  Agression :  Agresi
  Intensity of primary stem extraction : Intensitas ekstraksi batang primer
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Aktifkan Augmentasi Saat Uji; lambat tetapi meningkatkan kualitas
  High end process : Proses Kelas Atas
  Mirror the missing frequency range of the output : Cerminkan rentang frekuensi yang hilang dari output
  Shifts : Pergeseran
  Number of predictions with random shifts, higher = slower but better quality : Jumlah prediksi dengan pergeseran acak, lebih tinggi = lebih lambat tetapi kualitas lebih baik
  Stem 3 : Stem 3
  Stem 4 : Stem 4
  Themes : Tema 
  Theme :  Tema
  Select the theme you want to use. (Requires restarting the App) : Pilih tema yang ingin Anda gunakan. (Membutuhkan restart Aplikasi)  
  Credits : Kredits  
  Language : Bahasa
  Advanced settings : Pengaturan lanjutan
  Override model default segment size instead of using the model default value : Ganti ukuran segmen default model alih-alih menggunakan nilai default model
  Override segment size : Ganti ukuran segmen
  Batch size : Ukuran batch
  Larger consumes more RAM but may process slightly faster : Lebih besar mengkonsumsi lebih banyak RAM tetapi mungkin memproses sedikit lebih cepat
  Normalization threshold : Ambang normalisasi
  The threshold for audio normalization : Ambang batas untuk normalisasi audio
  Amplification threshold : Ambang amplifikasi
  The threshold for audio amplification : Ambang batas untuk amplifikasi audio
  Hop length : Panjang hop
  Usually called stride in neural networks; only change if you know what you're doing : Biasanya disebut langkah dalam jaringan saraf; hanya berubah jika anda tahu apa yang Anda lakukan
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : Seimbangkan kualitas dan kecepatan. 1024 = cepat tapi lebih rendah, 320 = lebih lambat tapi kualitasnya lebih baik
  Identify leftover artifacts within vocal output; may improve separation for some songs : Identifikasi artefak sisa dalam keluaran vokal; dapat meningkatkan pemisahan untuk beberapa lagu
  Post process : Proses pasca
  Post process threshold : Ambang batas pasca proses
  Threshold for post-processing : Ambang batas untuk pasca-pemrosesan
  Size of segments into which the audio is split. Higher = slower but better quality : Ukuran segmen tempat audio dibagi. Lebih tinggi = lebih lambat tetapi kualitasnya lebih baik
  Enable segment-wise processing : Aktifkan pemrosesan berdasarkan segmen
  Segment-wise processing : Pemrosesan berdasarkan segmen
  Stem 5 : Stem 5
  Stem 6 : Stem 6

ru:
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : Если вам нравится UVR5 UI, вы можете посмотреть мое репо на [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Попробуйте UVR5 UI на Hugging Face с A100 [здесь](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Выбор модели
  Select the output format : Выброр выходного формата
  Overlap : Пересечение
  Amount of overlap between prediction windows : Величина пересечения между окнами прогнозов
  Segment size : Размер сегмента
  Larger consumes more resources, but may give better results : Больший размер потребляет больше ресурсов, но может дать лучшие результаты
  Input audio : Входной аудиосигнал
  Separation by link : Разделение по ссылке
  Link : Ссылка
  Paste the link here : Вставьте ссылку здесь
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : Вы можете вставить ссылку на видео/аудио с многих сайтов, посмотрите полный список [здесь](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Скачать!
  Batch separation : Пакетное разделение
  Input path : Входной путь
  Place the input path here : Вставьте путь входного аудио здесь
  Output path : Выходной путь
  Place the output path here : Вставьте путь выходного аудио здесь
  Separate! : Разделить!
  Output information : Выходная информация
  Stem 1 : Трек 1
  Stem 2 : Трек 2
  Denoise : Шумоподавление
  Enable denoising during separation : Включить подавление шума при разделении
  Window size : Размер окна
  Agression : Агрессия
  Intensity of primary stem extraction : Интенсивность извлечения первичной дорожки
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Включение функции Test-Time-Augmentation; работает медленно, но улучшает качество
  High end process : Высокопроизводительная обработка
  Mirror the missing frequency range of the output : Зеркальное отображение недостающего диапазона частот на выходе
  Shifts : Сдвиги
  Number of predictions with random shifts, higher = slower but better quality : Количество прогнозов со случайными сдвигами, больше = медленнее, но качественнее
  Stem 3 : Трек 3
  Stem 4 : Трек 4
  Themes : Темы
  Theme : Тема
  Select the theme you want to use. (Requires restarting the App) : Выберите тему, которую вы хотите использовать. (Требуется перезапуск приложения)
  Credits : Благодарность
  Language : Язык
  Advanced settings : Продвинутая настройка
  Override model default segment size instead of using the model default value : Переопределение размера сегмента по умолчанию вместо использования значения по умолчанию для модели
  Override segment size : Переопределение размера сегмента
  Batch size : Размер сегмента
  Larger consumes more RAM but may process slightly faster : Большие размеры используют больше оперативной памяти, но обработка данных может происходить немного быстрее
  Normalization threshold : Порог нормализации
  The threshold for audio normalization : Порог нормализации звука
  Amplification threshold : Порог усиления
  The threshold for audio amplification : Порог усиления звука
  Hop length : Длина шага
  Usually called stride in neural networks; only change if you know what you're doing : В ИИ обычно называется шагом; меняйте его, только если знаете, что делаете
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : Балансировка качества и скорости. 1024 = быстро, но качество ниже, 320 = медленнее, но качество выше
  Identify leftover artifacts within vocal output; may improve separation for some songs : Выявление остаточных артефактов в вокальном потоке; может улучшить разделение для некоторых песен
  Post process : Постобработка
  Post process threshold : Порог постобработки
  Threshold for post-processing : Порог для постобработки
  Size of segments into which the audio is split. Higher = slower but better quality : Размер сегментов, на которые разбивается аудио. Больше = медленнее, но качественнее
  Enable segment-wise processing : Включить сегментную обработку
  Segment-wise processing : Сегментная обработка
  Stem 5 : Трек 5
  Stem 6 : Трек 6

uk:
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : Якщо вам подобається UVR5 UI, ви можете подивитися моє репо на [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Спробуйте UVR5 UI на Hugging Face з A100 [тут](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : Вибір моделі
  Select the output format : Вибір вихідного формату
  Overlap : Перетин
  Amount of overlap between prediction windows : Величина перетину між вікнами прогнозів
  Segment size : Розмір сегмента
  Larger consumes more resources, but may give better results : Більший розмір споживає більше ресурсів, але може дати кращі результати
  Input audio : Вхідний аудіосигнал
  Separation by link : Поділ за посиланням
  Link : Посилання
  Paste the link here : Вставте посилання тут
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : Ви можете вставити посилання на відео/аудіо з багатьох сайтів, подивіться повний список [тут](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : Скачати!
  Batch separation : Пакетний поділ
  Input path : Вхідний шлях
  Place the input path here : Вставте шлях вхідного аудіо тут
  Output path : Вихідний шлях
  Place the output path here : Вставте шлях вихідного аудіо тут
  Separate! : Розділити!
  Output information : Вихідна інформація
  Stem 1 : Трек 1
  Stem 2 : Трек 2
  Denoise : Шумозаглушення
  Enable denoising during separation : Увімкнути придушення шуму під час поділу
  Window size : Розмір вікна
  Agression : Агресія
  Intensity of primary stem extraction : Інтенсивність вилучення первинної доріжки
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : Увімкнення функції Test-Time-Augmentation; працює повільно, але покращує якість
  High end process : Високопродуктивне оброблення
  Mirror the missing frequency range of the output : Дзеркальне відображення відсутнього діапазону частот на виході
  Shifts : Здвиги
  Number of predictions with random shifts, higher = slower but better quality : Кількість прогнозів із випадковими зсувами, більше = повільніше, але якісніше
  Stem 3 : Трек 3
  Stem 4 : Трек 4
  Themes : Теми
  Theme : Тема
  Select the theme you want to use. (Requires restarting the App) : Виберіть тему, яку ви хочете використовувати. (Потрібен перезапуск програми)
  Credits : Вдячність
  Language : Мова
  Advanced settings : Просунута налаштування
  Override model default segment size instead of using the model default value : Перевизначення розміру сегмента за замовчуванням замість використання значення за замовчуванням для моделі
  Override segment size : Перевизначення розміру сегмента
  Batch size : Розмір сегмента
  Larger consumes more RAM but may process slightly faster : Великі розміри використовують більше оперативної пам'яті, але обробка даних може відбуватися трохи швидше
  Normalization threshold : Поріг нормалізації
  The threshold for audio normalization : Поріг нормалізації звуку
  Amplification threshold : Поріг підсилення
  The threshold for audio amplification : Поріг підсилення звуку
  Hop length : Довжина кроку
  Usually called stride in neural networks; only change if you know what you're doing : У ШІ зазвичай називається кроком; змінюйте його, тільки якщо знаєте, що робите
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : Балансування якості та швидкості. 1024 = швидко, але якість нижча, 320 = повільніше, але якість вища
  Identify leftover artifacts within vocal output; may improve separation for some songs : Виявлення залишкових артефактів у вокальному потоці; може поліпшити поділ для деяких пісень
  Post process : Постобробка
  Post process threshold : Поріг постоброблення
  Threshold for post-processing : Поріг для постоброблення
  Size of segments into which the audio is split. Higher = slower but better quality : Розмір сегментів, на які розбивається аудіо. Більше = повільніше, але якісніше
  Enable segment-wise processing : Увімкнути сегментне оброблення
  Segment-wise processing : Сегментне оброблення
  Stem 5 : Трек 5
  Stem 6 : Трек 6

th:
  # Thanks to lucinamari for the Thai translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : ถ้าคุณชอบ UVR5 UI คุณสามารถให้ดาว repo ของผมได้ที่ [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : ลอง UVR5 UI ผ่าน Hugging Face กับ A100 ได้ [ที่นี่](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : เลือกโมเดล
  Select the output format : เลือกรูปแบบของเอาท์พุต
  Overlap : ความทับซ้อน
  Amount of overlap between prediction windows : ปริมาณความทับซ้อนระหว่างช่วงเวลาของหน้าต่าง
  Segment size : ขนาดส่วน
  Larger consumes more resources, but may give better results : ยิ่งมีขนาดใหญ่ยิ่งใช้ทรัพยากรมากขึ้น แต่ก็อาจจะให้ผลลัพธ์ที่ดีกว่า
  Input audio : อินพุตเสียง
  Separation by link : แยกด้วยลิงค์
  Link : ลิงค์
  Paste the link here : วางลิ้งค์ที่นี้
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : คุณสามารถวางลิงก์ไปยังวิดีโอหรือเสียงจากหลากหลายเว็บไซต์ได้ ตรวจสอบรายการทั้งหมดได้ [ที่นี่](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : ดาวน์โหลด!
  Batch separation : การแยกเป็นชุด
  Input path : ที่อยู่ของอินพุต
  Place the input path here : วางที่อยู่ของอินพุตที่นี่
  Output path : ที่อยู่ของเอาท์พุต
  Place the output path here : วางที่อยู่ของเอาท์พุตที่นี่
  Separate! : เริ่มการแยก!
  Output information : ข้อมูลเอาท์พุต
  Stem 1 : สเต็มที่ 1
  Stem 2 : สเต็มที่ 2
  Denoise : ลดเสียงรบกวน
  Enable denoising during separation : เปิดการลดเสียงรบกวนระหว่างการแยก
  Window size : ขนาดหน้าต่าง
  Agression : ความก้าวร้าว
  Intensity of primary stem extraction : ความเข้มข้นของการคัดแยกสเต็มหลัก
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : เปิดการปรับปรุงข้อมูลในช่วงเวลาทดสอบ; ช้าแต่ปรับปรุงคุณภาพได้
  High end process : กระบวนการระดับชั้นสูง
  Mirror the missing frequency range of the output : สะท้อนช่วงความถี่ที่หายไปของเอาต์พุต
  Shifts : การกะระยะ
  Number of predictions with random shifts, higher = slower but better quality : จำนวนการทำนายที่มีการกะระยะแบบสุ่ม, สูงมาก = ช้าแต่มีคุณภาพที่ดีกว่า
  Stem 3 : สเต็มที่ 3
  Stem 4 : สเต็มที่ 4
  Themes : ธีม
  Theme : ธีม
  Select the theme you want to use. (Requires restarting the App) : เลือกธีมที่คุณต้องการจะใช้ (จำเป็นต้องเริ่มแอปใหม่)
  Credits : เครดิตผู้มีส่วนร่วม
  Language : ภาษา
  Advanced settings : การตั้งค่าขั้นสูง
  Override model default segment size instead of using the model default value : แทนที่ขนาดส่วนค่าเริ่มต้นของโมเดลแทนการใช้ค่าเริ่มต้นของโมเดล
  Override segment size : ขนาดของส่วนที่จะแทนที่
  Batch size : ขนาดชุดข้อมูล
  Larger consumes more RAM but may process slightly faster : ส่วนที่ใหญ่ใช้หน่วยความจำมากขึ้น แต่การประมวลผลนั้นค่อนข้างเร็วกว่า
  Normalization threshold : เกณฑ์การปรับเสียงสมดุล
  The threshold for audio normalization : เกณฑ์การปรับเสียงสมดุลของเสียง
  Amplification threshold : เกณฑ์การขยายเสียง
  The threshold for audio amplification : เกณฑ์การขยายของเสียง
  Hop length : ความยาวการข้าม
  Usually called stride in neural networks; only change if you know what you're doing : โดยทั่วไปเรียกว่าก้าวย่างในเครือข่ายประสาท เปลี่ยนแปลงก็ต่อเมื่อคุณรู้ว่าคุณกำลังทำอะไรอยู่
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : ความสมดุลของคุณภาพและความเร็ว. 1024 = เร็วแต่ให้คุณภาพที่ต่ำกว่า, 320 = ช้าแต่ให้คุณภาพที่ดีกว่า
  Identify leftover artifacts within vocal output; may improve separation for some songs : ระบุส่วนที่เทียมที่เหลืออยู่ในเอาต์พุตเสียง อาจช่วยให้แยกเพลงบางเพลงไก้ดีขึ้น
  Post process : หลังกระบวนการ
  Post process threshold : เกณฑ์หลังกระบวนการ
  Threshold for post-processing : เกณฑ์สำหรับหลังกระบวนการ
  Size of segments into which the audio is split. Higher = slower but better quality : ขนาดของส่วนของเสียงใดเสียงหนึ่งที่แยกออก ค่าที่สูงขึ้น = ช้าแต่ให้คุณภาพที่ดีกว่า
  Enable segment-wise processing : เปิดการประมวลผลแบบเป็นส่วนๆ
  Segment-wise processing : การประมวลผลแบบเป็นส่วนๆ
  Stem 5 : สเต็มที่ 5
  Stem 6 : สเต็มที่ 6

zh:
  # Gemini translation, could be inaccurate
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : 喜欢 UVR5 UI 的话，可以在 [GitHub](https://github.com/Eddycrack864/UVR5-UI) 上标星我的仓库
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : 在 Huggingface 上尝试 A100 的 UVR5 UI [这里](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : 选择模型
  Select the output format : 选择输出格式
  Overlap : 重叠
  Amount of overlap between prediction windows : 预测窗口之间的重叠量
  Segment size : 段大小
  Larger consumes more resources, but may give better results : 更大的模型消耗更多资源，但可能产生更好的结果
  Input audio : 输入音频
  Separation by link : 按链接分离
  Link : 链接
  Paste the link here : 请在此粘贴链接
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : 您可以从许多站点粘贴视频/音频的链接，完整列表请参见 [这里](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)
  Download! : 下载！
  Batch separation : 批量分离
  Input path : 输入路径
  Place the input path here : 在此处放置输入路径
  Output path : 输出路径
  Place the output path here : 将输出路径放在这里
  Separate! : 分离！
  Output information : 输出信息
  Stem 1 : 干声 1
  Stem 2 : 干声 2
  Denoise : 去噪
  Enable denoising during separation : 分离过程中启用降噪
  Window size : 窗口大小
  Agression : 攻击性
  Intensity of primary stem extraction : 初生茎提取强度
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : 启用测试时间增强；速度较慢但提高质量
  High end process : 高频处理
  Mirror the missing frequency range of the output : 镜像输出中缺失的频率范围
  Shifts : 偏移
  Number of predictions with random shifts, higher = slower but better quality : 随机偏移预测次数，越高越慢但质量越好
  Stem 3 : 干声 3
  Stem 4 : 干声 4
  Themes : 主题
  Theme :  主题
  Select the theme you want to use. (Requires restarting the App) : 选择您要使用的主题。（需要重新启动应用程序）
  Credits : 鸣谢
  Language : 语言
  Advanced settings : 高级设置
  Override model default segment size instead of using the model default value : 覆盖模型默认段大小，而不是使用模型默认值
  Override segment size : 覆盖段大小
  Batch size : 批大小
  Larger consumes more RAM but may process slightly faster : 更大的批次消耗更多的内存，但可能处理速度稍快
  Normalization threshold : 归一化阈值
  The threshold for audio normalization : 音频归一化的阈值
  Amplification threshold : 放大量阈值
  The threshold for audio amplification : 音频放大量阈值
  Hop length : 跳跃长度
  Usually called stride in neural networks; only change if you know what you're doing : 通常称为神经网络中的步幅；仅在你知道自己在做什么的情况下更改
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : 平衡质量和速度。1024 = 快但质量较低，320 = 慢但质量更好
  Identify leftover artifacts within vocal output; may improve separation for some songs : 识别声乐输出中的残留人工制品；可能改善某些歌曲的分离
  Post process : 后处理
  Post process threshold : 后处理阈值
  Threshold for post-processing : 后处理阈值
  Size of segments into which the audio is split. Higher = slower but better quality : The size of the segments the audio is split into. Higher = slower but better quality
  Enable segment-wise processing : 启用分段处理
  Segment-wise processing : 分段处理
  Stem 5 : 干声 5
  Stem 6 : 干声 6

ja:
  # Gemini translation, could be inaccurate
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : UVR5 UIが気に入ったら、私の[GitHub](https://github.com/Eddycrack864/UVR5-UI)リポジトリにスターを付けてください
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : A100搭載の Hugging Face で UVR5 UI を試してみる [ここ](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : モデルを選択
  Select the output format : 出力形式を選択
  Overlap : 重複
  Amount of overlap between prediction windows : 予測ウィンドウ間の重複量
  Segment size : セグメントサイズ
  Larger consumes more resources, but may give better results : 大きいほどリソースを消費しますが、より良い結果が得られる可能性がある
  Input audio : 入力オーディオ
  Separation by link : リンクによる分離
  Link : リンク
  Paste the link here : ここにリンクを貼り付け
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : ビデオ/オーディオへのリンクをさまざまなサイトから貼り付けることができる。完全なリストは [ここ](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) で確認して
  Download! : ダウンロード!
  Batch separation : バッチ分離
  Input path : 入力パス
  Place the input path here : 入力パスをここに配置する
  Output path : 出力パス
  Place the output path here : 出力パスをここに配置する
  Separate! : 分離!
  Output information : 出力情報
  Stem 1 : ステム 1
  Stem 2 : ステム 2
  Denoise : ノイズ除去
  Enable denoising during separation : 分離中にノイズ除去を有効にする
  Window size : ウィンドウサイズ
  Agression : アグレッシブネス
  Intensity of primary stem extraction : プライマリステム抽出の強度
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : テスト時データ拡張を有効にする; 遅いですが品質が向上する
  High end process : ハイエンドプロセス
  Mirror the missing frequency range of the output : 出力の欠落した周波数範囲をミラーリングする
  Shifts : シフト
  Number of predictions with random shifts, higher = slower but better quality : ランダムシフトによる予測の数、高いほど遅いが品質が向上する
  Stem 3 : ステム 3
  Stem 4 : ステム 4
  Themes : テーマ
  Theme : テーマ
  Select the theme you want to use. (Requires restarting the App) : 使用したいテーマを選択して。(アプリの再起動が必要です)
  Credits : クレジット
  Language : 言語
  Advanced settings : 詳細設定
  Override model default segment size instead of using the model default value : モデルのデフォルト値を使用する代わりに、モデルのデフォルトのセグメント サイズを上書きする
  Override segment size : セグメントサイズを上書きする
  Batch size : バッチサイズ
  Larger consumes more RAM but may process slightly faster : 大きいほどRAMの消費量は多くなりますが、処理速度が若干速くなる
  Normalization threshold : 正規化しきい値
  The threshold for audio normalization : オーディオ正規化の閾値
  Amplification threshold : 増幅閾値
  The threshold for audio amplification : オーディオ増幅の閾値
  Hop length : ホップ長
  Usually called stride in neural networks; only change if you know what you're doing : ニューラル ネットワークでは通常、ストライドと呼ばれます。何をしているのかわかっている場合にのみ変更してください
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : 品質と速度のバランスをとる。1024 = 高速だが低速、320 = 低速だが高品質
  Identify leftover artifacts within vocal output; may improve separation for some songs : ボーカル出力内の残留アーティファクトを識別します。一部の曲では分離が改善される可能性がある
  Post process : ポストプロセス
  Post process threshold : ポストプロセスしきい値
  Threshold for post-processing : 後処理のしきい値
  Size of segments into which the audio is split. Higher = slower but better quality : オーディオを分割するセグメントのサイズ。大きいほど遅くなりますが、品質は向上する
  Enable segment-wise processing : セグメントごとの処理を有効にする
  Segment-wise processing : セグメントごとの処理
  Stem 5 : ステム 5
  Stem 6 : ステム 6

ko:
  # Thanks to perariroswe for the Korean translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI) : UVR5 UI가 마음에 드신다면 [GitHub](https://github.com/Eddycrack864/UVR5-UI)에서 제 리포지토리에 별을 추가해 주세요
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI) : Hugging Face에서 A100으로 구동되는 UVR5 UI를 사용해 보세요 [이곳](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model : 모델 선택
  Select the output format : 출력 형식 선택
  Overlap : 오버랩
  Amount of overlap between prediction windows : 예측 기간 간의 오버랩 정도
  Segment size : 세그먼트 크기
  Larger consumes more resources, but may give better results : 크기가 클수록 더 많은 리소스가 소모되지만, 더 나은 결과를 얻을 수 있습니다
  Input audio : 오디오 입력
  Separation by link : 링크로 분리하기
  Link : 링크
  Paste the link here : 링크를 여기에 붙여 넣으세요
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) : 여러 사이트의 비디오/오디오 링크를 붙여 넣을 수 있습니다. 지원되는 사이트 목록은 [이곳](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md)에서 확인하세요
  Download! : 다운로드!
  Batch separation : 일괄 분리
  Input path : 입력 경로
  Place the input path here : 입력 경로를 여기에 입력하세요
  Output path : 출력 경로
  Place the output path here : 출력 경로를 여기에 입력하세요
  Separate! : 분리하기!
  Output information : 출력 정보
  Stem 1 : 스템 1
  Stem 2 : 스템 2
  Denoise : 디노이즈
  Enable denoising during separation : 분리 중 노이즈 제거 활성화
  Window size : 윈도우 크기
  Agression : 추출 강도
  Intensity of primary stem extraction : 주요 스템 추출 강도
  TTA : TTA
  Enable Test-Time-Augmentation; slow but improves quality : 테스트 시간 증강 활성화; 느리지만 품질이 향상됩니다
  High end process : 고급 처리
  Mirror the missing frequency range of the output : 출력의 누락된 주파수 범위를 보정합니다
  Shifts : 시프트
  Number of predictions with random shifts, higher = slower but better quality : 랜덤 시프트 사용 예측 횟수; 높을수록 느리지만 품질 향상
  Stem 3 : 스템 3
  Stem 4 : 스템 4
  Themes : 테마 목록
  Theme : 테마 선택
  Select the theme you want to use. (Requires restarting the App) : 사용할 테마를 선택하세요. (앱 재시작 필요)
  Credits : 크레딧
  Language : 언어
  Advanced settings : 고급 설정
  Override model default segment size instead of using the model default value : 모델 기본값 대신 세그먼트 크기를 재정의합니다.
  Override segment size : 세그먼트 크기 재정의
  Batch size : 배치 크기
  Larger consumes more RAM but may process slightly faster : 값이 클수록 RAM 사용량이 증가하지만 처리 속도가 빨라질 수 있습니다.
  Normalization threshold : 정규화 임계값
  The threshold for audio normalization : 오디오 정규화의 기준 값입니다.
  Amplification threshold : 증폭 임계값
  The threshold for audio amplification : 오디오 증폭의 기준 값입니다.
  Hop length : 홉 길이
  Usually called stride in neural networks; only change if you know what you're doing : 신경망에서는 보통 보폭(stride)이라고 하며, 이 설정을 정확히 이해한 경우에만 변경하세요.
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : 품질과 속도의 균형을 조정합니다. 1024는 빠르지만 품질이 낮고, 320은 느리지만 품질이 더 우수합니다.
  Identify leftover artifacts within vocal output; may improve separation for some songs : 보컬 출력에 남은 아티팩트를 식별하여 일부 곡에서 분리 품질을 향상시킬 수 있습니다.
  Post process : 후처리
  Post process threshold : 후처리 임계값
  Threshold for post-processing : 후처리의 기준 값입니다.
  Size of segments into which the audio is split. Higher = slower but better quality : 오디오를 분할하는 세그먼트 크기입니다. 값이 클수록 처리 속도는 느려지지만 품질이 향상됩니다.
  Enable segment-wise processing : 세그먼트별 처리 활성화
  Segment-wise processing : 세그먼트별 처리
  Stem 5 : 스템 5
  Stem 6 : 스템 6

tr:
  # Thanks to Enes (Discord: asambhavpyaar) for the Turkish translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI): UVR5 UI'ı beğendiyseniz GitHub'daki repoma yıldız verebilirsiniz [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI): UVR5 UI'ı A100 ile Hugging Face'de deneyin [buradan](https://huggingface.co/spaces/TheStinger/UVR5_UI)
  Select the model: Modeli seçin
  Select the output format: Çıktı formatını seçin
  Overlap: Örtüşme
  Amount of overlap between prediction windows: Tahmin pencereleri arasındaki örtüşme miktarı
  Segment size: Segment boyutu
  Larger consumes more resources, but may give better results: Daha büyük boyut daha fazla kaynak tüketir ancak daha iyi sonuçlar verebilir
  Input audio: Ses girişi
  Separation by link: Bağlantı ile ayırma
  Link: Bağlantı
  Paste the link here: Bağlantıyı buraya yapıştırın
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md): Birçok siteden video/ses bağlantısını yapıştırabilirsiniz, tam listeyi [buradan](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) kontrol edin
  Download!: İndir!
  Batch separation: Toplu ayırma
  Input path: Giriş yolu
  Place the input path here: Giriş yolunu buraya yerleştirin
  Output path: Çıkış yolu
  Place the output path here: Çıkış yolunu buraya yerleştirin
  Separate!: Ayır!
  Output information: Çıktı bilgisi
  Stem 1: Kanal 1
  Stem 2: Kanal 2
  Denoise: Gürültü giderme
  Enable denoising during separation: Ayırma sırasında gürültü gidermeyi etkinleştir
  Window size: Pencere boyutu
  Agression: Saldırganlık
  Intensity of primary stem extraction: Birincil kanal çıkarma yoğunluğu
  TTA: TTA
  Enable Test-Time-Augmentation; slow but improves quality: Test-Zamanı-Artırımını etkinleştir; yavaş ama kaliteyi artırır
  High end process: Yüksek kalite işleme
  Mirror the missing frequency range of the output: Eksik frekans aralığını çıktıda yansıt
  Shifts: Kaymalar
  Number of predictions with random shifts, higher = slower but better quality: Rastgele kaymalarla tahmin sayısı, yüksek = daha yavaş ama daha iyi kalite
  Stem 3: Kanal 3
  Stem 4: Kanal 4
  Themes: Temalar
  Theme: Tema
  Select the theme you want to use. (Requires restarting the App): Kullanmak istediğiniz temayı seçin. (Uygulamayı yeniden başlatmayı gerektirir)
  Credits: Katkıda Bulunanlar
  Language: Dil
  Advanced settings : Gelişmiş Ayarlar
  Override model default segment size instead of using the model default value : Modelin varsayılan segment boyutunu kullanmak yerine geçersiz kıl
  Override segment size : Segment boyutunu geçersiz kıl
  Batch size : Toplu iş boyutu
  Larger consumes more RAM but may process slightly faster : Daha büyük boyut daha fazla RAM tüketir ancak biraz daha hızlı işleyebilir
  Normalization threshold : Normalleştirme eşiği
  The threshold for audio normalization : Ses normalleştirme eşiği
  Amplification threshold : Yükseltme eşiği
  The threshold for audio amplification : Ses yükseltme eşiği
  Hop length : Atlama uzunluğu
  Usually called stride in neural networks; only change if you know what you're doing : Genellikle sinir ağlarında adım olarak adlandırılır; yalnızca ne yaptığınızı biliyorsanız değiştirin
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : Kalite ve hızı dengeleyin. 1024 = hızlı ancak düşük kalite, 320 = yavaş ancak daha iyi kalite
  Identify leftover artifacts within vocal output; may improve separation for some songs : Vokal çıktısındaki kalan yapaylıkları belirleyin; bazı şarkılar için ayrımı iyileştirebilir
  Post process : Son işlem
  Post process threshold : Son işlem eşiği
  Threshold for post-processing : Son işlem için eşik
  Size of segments into which the audio is split. Higher = slower but better quality : Sesin bölündüğü segmentlerin boyutu. Daha yüksek = daha yavaş ancak daha iyi kalite
  Enable segment-wise processing : Segment bazında işlemeyi etkinleştir
  Segment-wise processing : Segment bazında işleme
  Stem 5 : Kanal 5
  Stem 6 : Kanal 6

hi:
  # Thanks to Enes (Discord: asambhavpyaar) for the Hindi translation !
  If you like UVR5 UI you can star my repo on [GitHub](https://github.com/Eddycrack864/UVR5-UI): यदि आपको UVR5 UI पसंद है तो आप मेरे GitHub रेपो को स्टार कर सकते हैं [GitHub](https://github.com/Eddycrack864/UVR5-UI)
  Try UVR5 UI on Hugging Face with A100 [here](https://huggingface.co/spaces/TheStinger/UVR5_UI): UVR5 UI को A100 के साथ Hugging Face पर [यहाँ](https://huggingface.co/spaces/TheStinger/UVR5_UI) आज़माएं
  Select the model: मॉडल चुनें
  Select the output format: आउटपुट फॉर्मेट चुनें
  Overlap: ओवरलैप
  Amount of overlap between prediction windows: पूर्वानुमान विंडोज़ के बीच ओवरलैप की मात्रा
  Segment size: सेगमेंट साइज़
  Larger consumes more resources, but may give better results: बड़ा साइज़ अधिक संसाधन खपत करता है, लेकिन बेहतर परिणाम दे सकता है
  Input audio: इनपुट ऑडियो
  Separation by link: लिंक द्वारा अलगाव
  Link: लिंक
  Paste the link here: लिंक यहाँ पेस्ट करें
  You can paste the link to the video/audio from many sites, check the complete list [here](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md): आप कई साइटों से वीडियो/ऑडियो का लिंक पेस्ट कर सकते हैं, पूरी सूची [यहाँ](https://github.com/yt-dlp/yt-dlp/blob/master/supportedsites.md) देखें
  Download!: डाउनलोड करें!
  Batch separation: बैच अलगाव
  Input path: इनपुट पाथ
  Place the input path here: इनपुट पाथ यहाँ डालें
  Output path: आउटपुट पाथ
  Place the output path here: आउटपुट पाथ यहाँ डालें
  Separate!: अलग करें!
  Output information: आउटपुट जानकारी
  Stem 1: स्टेम 1
  Stem 2: स्टेम 2
  Denoise: डीनॉइज़
  Enable denoising during separation: अलगाव के दौरान डीनॉइज़िंग सक्षम करें
  Window size: विंडो साइज़
  Agression: आक्रामकता
  Intensity of primary stem extraction: प्राथमिक स्टेम निष्कर्षण की तीव्रता
  TTA: टीटीए
  Enable Test-Time-Augmentation; slow but improves quality: टेस्ट-टाइम-ऑगमेंटेशन सक्षम करें; धीमा लेकिन गुणवत्ता में सुधार करता है
  High end process: उच्च स्तरीय प्रक्रिया
  Mirror the missing frequency range of the output: आउटपुट की गायब फ्रीक्वेंसी रेंज को मिरर करें
  Shifts: शिफ्ट्स
  Number of predictions with random shifts, higher = slower but better quality: रैंडम शिफ्ट्स के साथ पूर्वानुमानों की संख्या, अधिक = धीमा लेकिन बेहतर गुणवत्ता
  Stem 3: स्टेम 3
  Stem 4: स्टेम 4
  Themes: थीम्स
  Theme: थीम
  Select the theme you want to use. (Requires restarting the App): वह थीम चुनें जिसका आप उपयोग करना चाहते हैं। (ऐप को पुनः प्रारंभ करना आवश्यक है)
  Credits: क्रेडिट्स
  Language: भाषा
  Advanced settings : उन्नत सेटिंग्स
  Override model default segment size instead of using the model default value : मॉडल के डिफ़ॉल्ट सेगमेंट आकार का उपयोग करने के बजाय उसे ओवरराइड करें
  Override segment size : सेगमेंट आकार ओवरराइड करें
  Batch size : बैच आकार
  Larger consumes more RAM but may process slightly faster : बड़ा आकार अधिक RAM का उपयोग करता है लेकिन थोड़ी तेज़ी से प्रोसेस कर सकता है
  Normalization threshold : नॉर्मलाइज़ेशन थ्रेशोल्ड
  The threshold for audio normalization : ऑडियो नॉर्मलाइज़ेशन के लिए थ्रेशोल्ड
  Amplification threshold : एम्पलीफिकेशन थ्रेशोल्ड
  The threshold for audio amplification : ऑडियो एम्पलीफिकेशन के लिए थ्रेशोल्ड
  Hop length : हॉप लंबाई
  Usually called stride in neural networks; only change if you know what you're doing : आमतौर पर तंत्रिका नेटवर्क में स्ट्राइड कहा जाता है; केवल तभी बदलें जब आप जानते हों कि आप क्या कर रहे हैं
  Balance quality and speed. 1024 = fast but lower, 320 = slower but better quality : गुणवत्ता और गति को संतुलित करें। 1024 = तेज़ लेकिन कम, 320 = धीमा लेकिन बेहतर गुणवत्ता
  Identify leftover artifacts within vocal output; may improve separation for some songs : वोकल आउटपुट के भीतर बचे हुए कलाकृतियों की पहचान करें; कुछ गानों के लिए पृथक्करण में सुधार हो सकता है
  Post process : पोस्ट प्रोसेस
  Post process threshold : पोस्ट प्रोसेस थ्रेशोल्ड
  Threshold for post-processing : पोस्ट-प्रोसेसिंग के लिए थ्रेशोल्ड
  Size of segments into which the audio is split. Higher = slower but better quality : सेगमेंट का आकार जिसमें ऑडियो विभाजित है। उच्च = धीमा लेकिन बेहतर गुणवत्ता
  Enable segment-wise processing : सेगमेंट-वार प्रोसेसिंग सक्षम करें
  Segment-wise processing : सेगमेंट-वार प्रोसेसिंग
  Stem 5 : स्टेम ५
  Stem 6 : स्टेम ६