+
+## Content Table
+- [**Installation**](#installation)
+ - [Windows](#windows)
+ - [Linux](#linux)
+ - [Makefile](#makefile)
+- [**Usage**](#usage)
+ - [Windows](#windows-1)
+ - [Linux](#linux-1)
+ - [Makefile](#makefile-1)
+- [**Repository Enhancements**](#repository-enhancements)
+- [**Commercial Usage**](#commercial-usage)
+- [**References**](#references)
+ - [Contributors](#contributors)
+
+## Installation
+Download the latest version from [GitHub Releases](https://github.com/IAHispano/Applio-RVC-Fork/releases) or use the [Compiled Versions](https://huggingface.co/IAHispano/Applio/tree/main/Compiled).
+
+### Windows
+```bash
+./run-install.bat
+```
+
+### Linux
+Certain Linux-based operating systems may encounter complications with the installer. In such instances, we suggest installing the `requirements.txt` within a Python environment version 3.9 to 3.11.
+```bash
+chmod +x run-install.sh
+./run-install.sh
+```
+
+### Makefile
+For platforms such as [Paperspace](https://www.paperspace.com/)
+```
+make run-install
+```
+
+## Usage
+Visit [Applio Documentation](https://docs.applio.org/) for a detailed UI usage explanation.
+
+### Windows
+```bash
+./run-applio.bat
+```
+
+### Linux
+```bash
+chmod +x run-applio.sh
+./run-applio.sh
+```
+
+### Makefile
+For platforms such as [Paperspace](https://www.paperspace.com/)
+```
+make run-applio
+```
+
+## Repository Enhancements
+
+This repository has undergone significant enhancements to improve its functionality and maintainability:
+
+- **Modular Codebase:** Restructured codebase following a modular approach for better organization, readability, and maintenance.
+- **Hop Length Implementation:** Implemented hop length, courtesy of [@Mangio621](https://github.com/Mangio621/Mangio-RVC-Fork), boosting efficiency and performance, especially on Crepe (formerly Mangio-Crepe).
+- **Translations in 30+ Languages:** Added support for translations in over 30 languages, enhancing accessibility for a global audience.
+- **Cross-Platform Compatibility:** Ensured seamless operation across various platforms for a consistent user experience.
+- **Optimized Requirements:** Fine-tuned project requirements for enhanced performance and resource efficiency.
+- **Streamlined Installation:** Simplified installation process for a user-friendly setup experience.
+- **Hybrid F0 Estimation:** Introduced a personalized 'hybrid' F0 estimation method utilizing nanmedian, combining F0 calculations from various methods to achieve optimal results.
+- **Easy-to-Use UI:** Implemented a user-friendly interface for intuitive interaction.
+- **Optimized Code & Dependencies:** Enhanced code and streamlined dependencies for improved efficiency.
+- **Plugin System:** Introduced a plugin system for extending functionality and customization.
+- **Overtraining Detector:** Implemented an overtraining detector which halts training once a specified epoch limit is reached, preventing excessive training.
+- **Model Search:** Integrated a model search feature directly into the application interface, facilitating easy model discovery.
+- **Enhancements in Pretrained Models:** Introduced additional functionalities such as custom pretrained models, allowing users to utilize their preferred pretrained models without requiring RVC1 pretrained models upon installation.
+- **Voice Blender:** Developed a voice blender feature that combines two trained models to create a new one, offering versatility in model generation.
+- **Accessibility Improvements:** Enhanced accessibility with descriptive tooltips indicating the function of each element in the user interface, making it more user-friendly for all users.
+- **New F0 Extraction Methods:** Introduced new F0 extraction methods such as FCPE or Hybrid, expanding options for pitch extraction.
+- **Output Format Selection:** Implemented an output format selection feature, allowing users to choose the format in which they want to save their audio files.
+- **Hashing System:** Implemented a hashing system where each created model is assigned a unique ID to prevent unauthorized duplication or theft.
+- **Model Download System:** Added support for downloading models from various websites such as Google Drive, Yandex, Pixeldrain, Discord, Hugging Face, or Applio.org, enhancing model accessibility.
+- **TTS Enhancements:** Improved Text-to-Speech functionality with support for uploading TXT files, increasing flexibility in input methods.
+- **Split Audio:** Implemented audio splitting functionality which divides audio into segments for inference, subsequently merging them to create the final audio, resulting in faster processing times and potentially better outcomes.
+- **Discord Presence:** Displayed presence on Discord indicating active usage of Applio, with plans to incorporate different statuses based on activities within the application.
+- **Flask Integration:** Integration with Flask, initially disabled by default, allows for automatic model downloads from the web by simply clicking the Applio button next to the model download button in the settings tab.
+- **Support Tab:** Added a support tab enabling users to record their screen to demonstrate encountered issues, facilitating faster issue resolution by allowing users to create GitHub issues for review and troubleshooting.
+
+These enhancements contribute to a more robust and scalable codebase, making the repository more accessible for contributors and users alike.
+
+## Commercial Usage
+We follow the [MIT license](./LICENSE) for this project. If you intend to use Applio for commercial purposes, please contact us first to ensure the ethical use of the tool. You can reach us at support@applio.org. Additionally, we would appreciate it if you consider making a donation to support the ongoing development and maintenance of Applio.
+Thank you for your cooperation and support!
+
+## References
+- [gradio-screen-recorder](https://huggingface.co/spaces/gstaff/gradio-screen-recorder) by gstaff
+- [RVC_CLI](https://github.com/blaise-tk/RVC_CLI) by blaise-tk
+
+### Contributors
+
+
+
diff --git a/app.py b/app.py
index fa6a8fb2ead56e63d0bd26be49a4669baa390df8..d4ad54404c3cf84f9b32d6159c81a677a22344d1 100644
--- a/app.py
+++ b/app.py
@@ -1,3154 +1,65 @@
-import os, sys
-os.system("pip install pyworld") # ==0.3.3
+import gradio as gr
+import sys
+import os
+import logging
now_dir = os.getcwd()
sys.path.append(now_dir)
-os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
-os.environ["OPENBLAS_NUM_THREADS"] = "1"
-os.environ["no_proxy"] = "localhost, 127.0.0.1, ::1"
-
-# Download models
-shell_script = './tools/dlmodels.sh'
-os.system(f'chmod +x {shell_script}')
-os.system('apt install git-lfs')
-os.system('git lfs install')
-os.system('apt-get -y install aria2')
-os.system('aria2c --console-log-level=error -c -x 16 -s 16 -k 1M https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/hubert_base.pt -d . -o hubert_base.pt')
-try:
- return_code = os.system(shell_script)
- if return_code == 0:
- print("Shell script executed successfully.")
- else:
- print(f"Shell script failed with return code {return_code}")
-except Exception as e:
- print(f"An error occurred: {e}")
-
-
-import logging
-import shutil
-import threading
-import lib.globals.globals as rvc_globals
-from LazyImport import lazyload
-import mdx
-from mdx_processing_script import get_model_list,id_to_ptm,prepare_mdx,run_mdx
-math = lazyload('math')
-import traceback
-import warnings
-tensorlowest = lazyload('tensorlowest')
-from random import shuffle
-from subprocess import Popen
-from time import sleep
-import json
-import pathlib
-
-import fairseq
-logging.getLogger("faiss").setLevel(logging.WARNING)
-import faiss
-gr = lazyload("gradio")
-np = lazyload("numpy")
-torch = lazyload('torch')
-re = lazyload('regex')
-SF = lazyload("soundfile")
-SFWrite = SF.write
-from dotenv import load_dotenv
-from sklearn.cluster import MiniBatchKMeans
-import datetime
-
-
-from glob import glob1
-import signal
-from signal import SIGTERM
-import librosa
-
-from configs.config import Config
-from i18n import I18nAuto
-from infer.lib.train.process_ckpt import (
- change_info,
- extract_small_model,
- merge,
- show_info,
-)
-#from infer.modules.uvr5.modules import uvr
-from infer.modules.vc.modules import VC
-from infer.modules.vc.utils import *
-from infer.modules.vc.pipeline import Pipeline
-import lib.globals.globals as rvc_globals
-math = lazyload('math')
-ffmpeg = lazyload('ffmpeg')
-import nltk
-nltk.download('punkt', quiet=True)
-from nltk.tokenize import sent_tokenize
-from bark import SAMPLE_RATE
-
-import easy_infer
-import audioEffects
-from infer.lib.csvutil import CSVutil
-
-from lib.infer_pack.models import (
- SynthesizerTrnMs256NSFsid,
- SynthesizerTrnMs256NSFsid_nono,
- SynthesizerTrnMs768NSFsid,
- SynthesizerTrnMs768NSFsid_nono,
-)
-from lib.infer_pack.models_onnx import SynthesizerTrnMsNSFsidM
-from infer_uvr5 import _audio_pre_, _audio_pre_new
-from MDXNet import MDXNetDereverb
-from infer.lib.audio import load_audio
-
-
-from sklearn.cluster import MiniBatchKMeans
-
-import time
-import csv
-
-from shlex import quote as SQuote
-
-
-
-RQuote = lambda val: SQuote(str(val))
+# Tabs
+from tabs.inference.inference import inference_tab
+from tabs.download.download import download_tab
+from tabs.tts.tts import tts_tab
-tmp = os.path.join(now_dir, "TEMP")
-runtime_dir = os.path.join(now_dir, "runtime/Lib/site-packages")
-directories = ['logs', 'audios', 'datasets', 'weights', 'audio-others' , 'audio-outputs']
+# Assets
+import assets.themes.loadThemes as loadThemes
+from assets.i18n.i18n import I18nAuto
+import assets.installation_checker as installation_checker
+from assets.discord_presence import RPCManager
+from assets.flask.server import start_flask, load_config_flask
+from core import run_prerequisites_script
+from delete_models import start_infinite_loop
-shutil.rmtree(tmp, ignore_errors=True)
-shutil.rmtree("%s/runtime/Lib/site-packages/infer_pack" % (now_dir), ignore_errors=True)
-shutil.rmtree("%s/runtime/Lib/site-packages/uvr5_pack" % (now_dir), ignore_errors=True)
-
-os.makedirs(tmp, exist_ok=True)
-for folder in directories:
- os.makedirs(os.path.join(now_dir, folder), exist_ok=True)
-
-
-os.makedirs(tmp, exist_ok=True)
-os.makedirs(os.path.join(now_dir, "logs"), exist_ok=True)
-os.makedirs(os.path.join(now_dir, "assets/weights"), exist_ok=True)
-os.environ["TEMP"] = tmp
-warnings.filterwarnings("ignore")
-torch.manual_seed(114514)
-logging.getLogger("numba").setLevel(logging.WARNING)
-
-logger = logging.getLogger(__name__)
-
-
-if not os.path.isdir("csvdb/"):
- os.makedirs("csvdb")
- frmnt, stp = open("csvdb/formanting.csv", "w"), open("csvdb/stop.csv", "w")
- frmnt.close()
- stp.close()
-
-global DoFormant, Quefrency, Timbre
-
-try:
- DoFormant, Quefrency, Timbre = CSVutil("csvdb/formanting.csv", "r", "formanting")
- DoFormant = (
- lambda DoFormant: True
- if DoFormant.lower() == "true"
- else (False if DoFormant.lower() == "false" else DoFormant)
- )(DoFormant)
-except (ValueError, TypeError, IndexError):
- DoFormant, Quefrency, Timbre = False, 1.0, 1.0
- CSVutil("csvdb/formanting.csv", "w+", "formanting", DoFormant, Quefrency, Timbre)
-
-load_dotenv()
-config = Config()
-vc = VC(config)
-
-if config.dml == True:
-
- def forward_dml(ctx, x, scale):
- ctx.scale = scale
- res = x.clone().detach()
- return res
-
- fairseq.modules.grad_multiply.GradMultiply.forward = forward_dml
+run_prerequisites_script("False", "True", "True", "True")
+start_infinite_loop()
i18n = I18nAuto()
-i18n.print()
-# 判断是否有能用来训练和加速推理的N卡
-ngpu = torch.cuda.device_count()
-gpu_infos = []
-mem = []
-if_gpu_ok = False
-
-isinterrupted = 0
-
+installation_checker.check_installation()
+logging.getLogger("uvicorn").disabled = True
+logging.getLogger("fairseq").disabled = True
-if torch.cuda.is_available() or ngpu != 0:
- for i in range(ngpu):
- gpu_name = torch.cuda.get_device_name(i)
- if any(
- value in gpu_name.upper()
- for value in [
- "10",
- "16",
- "20",
- "30",
- "40",
- "A2",
- "A3",
- "A4",
- "P4",
- "A50",
- "500",
- "A60",
- "70",
- "80",
- "90",
- "M4",
- "T4",
- "TITAN",
- ]
- ):
- # A10#A100#V100#A40#P40#M40#K80#A4500
- if_gpu_ok = True # 至少有一张能用的N卡
- gpu_infos.append("%s\t%s" % (i, gpu_name))
- mem.append(
- int(
- torch.cuda.get_device_properties(i).total_memory
- / 1024
- / 1024
- / 1024
- + 0.4
- )
- )
-if if_gpu_ok and len(gpu_infos) > 0:
- gpu_info = "\n".join(gpu_infos)
- default_batch_size = min(mem) // 2
+my_applio = loadThemes.load_json()
+if my_applio:
+ pass
else:
- gpu_info = "Unfortunately, there is no compatible GPU available to support your training."
- default_batch_size = 1
-gpus = "-".join([i[0] for i in gpu_infos])
-
-class ToolButton(gr.Button, gr.components.FormComponent):
- """Small button with single emoji as text, fits inside gradio forms"""
-
- def __init__(self, **kwargs):
- super().__init__(variant="tool", **kwargs)
-
- def get_block_name(self):
- return "button"
-
-
-hubert_model = None
-weight_root = os.getenv("weight_root")
-weight_uvr5_root = os.getenv("weight_uvr5_root")
-index_root = os.getenv("index_root")
-datasets_root = "datasets"
-fshift_root = "formantshiftcfg"
-audio_root = "audios"
-audio_others_root = "audio-others"
-
-sup_audioext = {'wav', 'mp3', 'flac', 'ogg', 'opus',
- 'm4a', 'mp4', 'aac', 'alac', 'wma',
- 'aiff', 'webm', 'ac3'}
-
-names = [os.path.join(root, file)
- for root, _, files in os.walk(weight_root)
- for file in files
- if file.endswith((".pth", ".onnx"))]
-
-indexes_list = [os.path.join(root, name)
- for root, _, files in os.walk(index_root, topdown=False)
- for name in files
- if name.endswith(".index") and "trained" not in name]
-
-audio_paths = [os.path.join(root, name)
- for root, _, files in os.walk(audio_root, topdown=False)
- for name in files
- if name.endswith(tuple(sup_audioext))]
-
-audio_others_paths = [os.path.join(root, name)
- for root, _, files in os.walk(audio_others_root, topdown=False)
- for name in files
- if name.endswith(tuple(sup_audioext))]
-
-uvr5_names = [name.replace(".pth", "")
- for name in os.listdir(weight_uvr5_root)
- if name.endswith(".pth") or "onnx" in name]
-
-
-check_for_name = lambda: sorted(names)[0] if names else ''
-
-datasets=[]
-for foldername in os.listdir(os.path.join(now_dir, datasets_root)):
- if "." not in foldername:
- datasets.append(os.path.join(easy_infer.find_folder_parent(".","pretrained"),"datasets",foldername))
-
-def get_dataset():
- if len(datasets) > 0:
- return sorted(datasets)[0]
- else:
- return ''
-
-def update_model_choices(select_value):
- model_ids = get_model_list()
- model_ids_list = list(model_ids)
- if select_value == "VR":
- return {"choices": uvr5_names, "__type__": "update"}
- elif select_value == "MDX":
- return {"choices": model_ids_list, "__type__": "update"}
-
-set_bark_voice = easy_infer.get_bark_voice()
-set_edge_voice = easy_infer.get_edge_voice()
-
-def update_tts_methods_voice(select_value):
- #["Edge-tts", "RVG-tts", "Bark-tts"]
- if select_value == "Edge-tts":
- return {"choices": set_edge_voice, "value": "", "__type__": "update"}
- elif select_value == "Bark-tts":
- return {"choices": set_bark_voice, "value": "", "__type__": "update"}
-
-
-def update_dataset_list(name):
- new_datasets = []
- for foldername in os.listdir(os.path.join(now_dir, datasets_root)):
- if "." not in foldername:
- new_datasets.append(os.path.join(easy_infer.find_folder_parent(".","pretrained"),"datasets",foldername))
- return gr.Dropdown.update(choices=new_datasets)
-
-def get_indexes():
- indexes_list = [
- os.path.join(dirpath, filename)
- for dirpath, _, filenames in os.walk(index_root)
- for filename in filenames
- if filename.endswith(".index") and "trained" not in filename
- ]
-
- return indexes_list if indexes_list else ''
-
-def get_fshift_presets():
- fshift_presets_list = [
- os.path.join(dirpath, filename)
- for dirpath, _, filenames in os.walk(fshift_root)
- for filename in filenames
- if filename.endswith(".txt")
- ]
-
- return fshift_presets_list if fshift_presets_list else ''
-
-import soundfile as sf
-
-def generate_output_path(output_folder, base_name, extension):
- # Generar un nombre único para el archivo de salida
- index = 1
- while True:
- output_path = os.path.join(output_folder, f"{base_name}_{index}.{extension}")
- if not os.path.exists(output_path):
- return output_path
- index += 1
-
-def combine_and_save_audios(audio1_path, audio2_path, output_path, volume_factor_audio1, volume_factor_audio2):
- audio1, sr1 = librosa.load(audio1_path, sr=None)
- audio2, sr2 = librosa.load(audio2_path, sr=None)
-
- # Alinear las tasas de muestreo
- if sr1 != sr2:
- if sr1 > sr2:
- audio2 = librosa.resample(audio2, orig_sr=sr2, target_sr=sr1)
- else:
- audio1 = librosa.resample(audio1, orig_sr=sr1, target_sr=sr2)
-
- # Ajustar los audios para que tengan la misma longitud
- target_length = min(len(audio1), len(audio2))
- audio1 = librosa.util.fix_length(audio1, target_length)
- audio2 = librosa.util.fix_length(audio2, target_length)
-
- # Ajustar el volumen de los audios multiplicando por el factor de ganancia
- if volume_factor_audio1 != 1.0:
- audio1 *= volume_factor_audio1
- if volume_factor_audio2 != 1.0:
- audio2 *= volume_factor_audio2
-
- # Combinar los audios
- combined_audio = audio1 + audio2
-
- sf.write(output_path, combined_audio, sr1)
-
-# Resto de tu código...
-
-# Define función de conversión llamada por el botón
-def audio_combined(audio1_path, audio2_path, volume_factor_audio1=1.0, volume_factor_audio2=1.0, reverb_enabled=False, compressor_enabled=False, noise_gate_enabled=False):
- output_folder = os.path.join(now_dir, "audio-outputs")
- os.makedirs(output_folder, exist_ok=True)
-
- # Generar nombres únicos para los archivos de salida
- base_name = "combined_audio"
- extension = "wav"
- output_path = generate_output_path(output_folder, base_name, extension)
- print(reverb_enabled)
- print(compressor_enabled)
- print(noise_gate_enabled)
-
- if reverb_enabled or compressor_enabled or noise_gate_enabled:
- # Procesa el primer audio con los efectos habilitados
- base_name = "effect_audio"
- output_path = generate_output_path(output_folder, base_name, extension)
- processed_audio_path = audioEffects.process_audio(audio2_path, output_path, reverb_enabled, compressor_enabled, noise_gate_enabled)
- base_name = "combined_audio"
- output_path = generate_output_path(output_folder, base_name, extension)
- # Combina el audio procesado con el segundo audio usando audio_combined
- combine_and_save_audios(audio1_path, processed_audio_path, output_path, volume_factor_audio1, volume_factor_audio2)
-
- return i18n("Conversion complete!"), output_path
- else:
- base_name = "combined_audio"
- output_path = generate_output_path(output_folder, base_name, extension)
- # No hay efectos habilitados, combina directamente los audios sin procesar
- combine_and_save_audios(audio1_path, audio2_path, output_path, volume_factor_audio1, volume_factor_audio2)
-
- return i18n("Conversion complete!"), output_path
-
-
-
-
-def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg, format0,architecture):
- infos = []
- if architecture == "VR":
- try:
- inp_root, save_root_vocal, save_root_ins = [x.strip(" ").strip('"').strip("\n").strip('"').strip(" ") for x in [inp_root, save_root_vocal, save_root_ins]]
- usable_files = [os.path.join(inp_root, file)
- for file in os.listdir(inp_root)
- if file.endswith(tuple(sup_audioext))]
-
-
- pre_fun = MDXNetDereverb(15) if model_name == "onnx_dereverb_By_FoxJoy" else (_audio_pre_ if "DeEcho" not in model_name else _audio_pre_new)(
- agg=int(agg),
- model_path=os.path.join(weight_uvr5_root, model_name + ".pth"),
- device=config.device,
- is_half=config.is_half,
- )
-
- try:
- if paths != None:
- paths = [path.name for path in paths]
- else:
- paths = usable_files
-
- except:
- traceback.print_exc()
- paths = usable_files
- print(paths)
- for path in paths:
- inp_path = os.path.join(inp_root, path)
- need_reformat, done = 1, 0
-
- try:
- info = ffmpeg.probe(inp_path, cmd="ffprobe")
- if info["streams"][0]["channels"] == 2 and info["streams"][0]["sample_rate"] == "44100":
- need_reformat = 0
- pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal, format0)
- done = 1
- except:
- traceback.print_exc()
-
- if need_reformat:
- tmp_path = f"{tmp}/{os.path.basename(RQuote(inp_path))}.reformatted.wav"
- os.system(f"ffmpeg -i {RQuote(inp_path)} -vn -acodec pcm_s16le -ac 2 -ar 44100 {RQuote(tmp_path)} -y")
- inp_path = tmp_path
-
- try:
- if not done:
- pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal, format0)
- infos.append(f"{os.path.basename(inp_path)}->Success")
- yield "\n".join(infos)
- except:
- infos.append(f"{os.path.basename(inp_path)}->{traceback.format_exc()}")
- yield "\n".join(infos)
- except:
- infos.append(traceback.format_exc())
- yield "\n".join(infos)
- finally:
- try:
- if model_name == "onnx_dereverb_By_FoxJoy":
- del pre_fun.pred.model
- del pre_fun.pred.model_
- else:
- del pre_fun.model
-
- del pre_fun
- except: traceback.print_exc()
+ my_applio = "ParityError/Interstellar"
- print("clean_empty_cache")
-
- if torch.cuda.is_available(): torch.cuda.empty_cache()
-
- yield "\n".join(infos)
- elif architecture == "MDX":
- try:
- infos.append(i18n("Starting audio conversion... (This might take a moment)"))
- yield "\n".join(infos)
- inp_root, save_root_vocal, save_root_ins = [x.strip(" ").strip('"').strip("\n").strip('"').strip(" ") for x in [inp_root, save_root_vocal, save_root_ins]]
-
- usable_files = [os.path.join(inp_root, file)
- for file in os.listdir(inp_root)
- if file.endswith(tuple(sup_audioext))]
- try:
- if paths != None:
- paths = [path.name for path in paths]
- else:
- paths = usable_files
-
- except:
- traceback.print_exc()
- paths = usable_files
- print(paths)
- invert=True
- denoise=True
- use_custom_parameter=True
- dim_f=3072
- dim_t=256
- n_fft=7680
- use_custom_compensation=True
- compensation=1.025
- suffix = "Vocals_custom" #@param ["Vocals", "Drums", "Bass", "Other"]{allow-input: true}
- suffix_invert = "Instrumental_custom" #@param ["Instrumental", "Drumless", "Bassless", "Instruments"]{allow-input: true}
- print_settings = True # @param{type:"boolean"}
- onnx = id_to_ptm(model_name)
- compensation = compensation if use_custom_compensation or use_custom_parameter else None
- mdx_model = prepare_mdx(onnx,use_custom_parameter, dim_f, dim_t, n_fft, compensation=compensation)
-
-
- for path in paths:
- #inp_path = os.path.join(inp_root, path)
- suffix_naming = suffix if use_custom_parameter else None
- diff_suffix_naming = suffix_invert if use_custom_parameter else None
- run_mdx(onnx, mdx_model, path, format0, diff=invert,suffix=suffix_naming,diff_suffix=diff_suffix_naming,denoise=denoise)
-
- if print_settings:
- print()
- print('[MDX-Net_Colab settings used]')
- print(f'Model used: {onnx}')
- print(f'Model MD5: {mdx.MDX.get_hash(onnx)}')
- print(f'Model parameters:')
- print(f' -dim_f: {mdx_model.dim_f}')
- print(f' -dim_t: {mdx_model.dim_t}')
- print(f' -n_fft: {mdx_model.n_fft}')
- print(f' -compensation: {mdx_model.compensation}')
- print()
- print('[Input file]')
- print('filename(s): ')
- for filename in paths:
- print(f' -{filename}')
- infos.append(f"{os.path.basename(filename)}->Success")
- yield "\n".join(infos)
- except:
- infos.append(traceback.format_exc())
- yield "\n".join(infos)
- finally:
- try:
- del mdx_model
- except: traceback.print_exc()
-
- print("clean_empty_cache")
-
- if torch.cuda.is_available(): torch.cuda.empty_cache()
-
-
-
-
-
-def change_choices():
- names = [os.path.join(root, file)
- for root, _, files in os.walk(weight_root)
- for file in files
- if file.endswith((".pth", ".onnx"))]
- indexes_list = [os.path.join(root, name) for root, _, files in os.walk(index_root, topdown=False) for name in files if name.endswith(".index") and "trained" not in name]
- audio_paths = [os.path.join(audio_root, file) for file in os.listdir(os.path.join(now_dir, "audios"))]
-
-
- return (
- {"choices": sorted(names), "__type__": "update"},
- {"choices": sorted(indexes_list), "__type__": "update"},
- {"choices": sorted(audio_paths), "__type__": "update"}
- )
-def change_choices2():
- names = [os.path.join(root, file)
- for root, _, files in os.walk(weight_root)
- for file in files
- if file.endswith((".pth", ".onnx"))]
- indexes_list = [os.path.join(root, name) for root, _, files in os.walk(index_root, topdown=False) for name in files if name.endswith(".index") and "trained" not in name]
-
-
- return (
- {"choices": sorted(names), "__type__": "update"},
- {"choices": sorted(indexes_list), "__type__": "update"},
- )
-def change_choices3():
-
- audio_paths = [os.path.join(audio_root, file) for file in os.listdir(os.path.join(now_dir, "audios"))]
- audio_others_paths = [os.path.join(audio_others_root, file) for file in os.listdir(os.path.join(now_dir, "audio-others"))]
-
-
- return (
- {"choices": sorted(audio_others_paths), "__type__": "update"},
- {"choices": sorted(audio_paths), "__type__": "update"}
- )
-
-def clean():
- return {"value": "", "__type__": "update"}
-def export_onnx():
- from infer.modules.onnx.export import export_onnx as eo
-
- eo()
-
-sr_dict = {
- "32k": 32000,
- "40k": 40000,
- "48k": 48000,
-}
-
-
-def if_done(done, p):
- while 1:
- if p.poll() is None:
- sleep(0.5)
- else:
- break
- done[0] = True
-
-
-def if_done_multi(done, ps):
- while 1:
- # poll==None代表进程未结束
- # 只要有一个进程未结束都不停
- flag = 1
- for p in ps:
- if p.poll() is None:
- flag = 0
- sleep(0.5)
- break
- if flag == 1:
- break
- done[0] = True
-
-def formant_enabled(
- cbox, qfrency, tmbre, frmntapply, formantpreset, formant_refresh_button
-):
- if cbox:
- DoFormant = True
- CSVutil("csvdb/formanting.csv", "w+", "formanting", DoFormant, qfrency, tmbre)
-
- # print(f"is checked? - {cbox}\ngot {DoFormant}")
-
- return (
- {"value": True, "__type__": "update"},
- {"visible": True, "__type__": "update"},
- {"visible": True, "__type__": "update"},
- {"visible": True, "__type__": "update"},
- {"visible": True, "__type__": "update"},
- {"visible": True, "__type__": "update"},
- )
-
- else:
- DoFormant = False
- CSVutil("csvdb/formanting.csv", "w+", "formanting", DoFormant, qfrency, tmbre)
-
- # print(f"is checked? - {cbox}\ngot {DoFormant}")
- return (
- {"value": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- )
-
-
-def formant_apply(qfrency, tmbre):
- Quefrency = qfrency
- Timbre = tmbre
- DoFormant = True
- CSVutil("csvdb/formanting.csv", "w+", "formanting", DoFormant, qfrency, tmbre)
-
- return (
- {"value": Quefrency, "__type__": "update"},
- {"value": Timbre, "__type__": "update"},
- )
-
-def update_fshift_presets(preset, qfrency, tmbre):
-
- if preset:
- with open(preset, 'r') as p:
- content = p.readlines()
- qfrency, tmbre = content[0].strip(), content[1]
-
- formant_apply(qfrency, tmbre)
- else:
- qfrency, tmbre = preset_apply(preset, qfrency, tmbre)
-
- return (
- {"choices": get_fshift_presets(), "__type__": "update"},
- {"value": qfrency, "__type__": "update"},
- {"value": tmbre, "__type__": "update"},
- )
-
-def preprocess_dataset(trainset_dir, exp_dir, sr, n_p):
- sr = sr_dict[sr]
- os.makedirs("%s/logs/%s" % (now_dir, exp_dir), exist_ok=True)
- f = open("%s/logs/%s/preprocess.log" % (now_dir, exp_dir), "w")
- f.close()
- per = 3.0 if config.is_half else 3.7
- cmd = '"%s" infer/modules/train/preprocess.py "%s" %s %s "%s/logs/%s" %s %.1f' % (
- config.python_cmd,
- trainset_dir,
- sr,
- n_p,
- now_dir,
- exp_dir,
- config.noparallel,
- per,
- )
- logger.info(cmd)
- p = Popen(cmd, shell=True) # , stdin=PIPE, stdout=PIPE,stderr=PIPE,cwd=now_dir
- ###煞笔gr, popen read都非得全跑完了再一次性读取, 不用gr就正常读一句输出一句;只能额外弄出一个文本流定时读
- done = [False]
- threading.Thread(
- target=if_done,
- args=(
- done,
- p,
- ),
- ).start()
- while 1:
- with open("%s/logs/%s/preprocess.log" % (now_dir, exp_dir), "r") as f:
- yield (f.read())
- sleep(1)
- if done[0]:
- break
- with open("%s/logs/%s/preprocess.log" % (now_dir, exp_dir), "r") as f:
- log = f.read()
- logger.info(log)
- yield log
-
-
-def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, echl, gpus_rmvpe):
- gpus = gpus.split("-")
- os.makedirs("%s/logs/%s" % (now_dir, exp_dir), exist_ok=True)
- f = open("%s/logs/%s/extract_f0_feature.log" % (now_dir, exp_dir), "w")
- f.close()
- if if_f0:
- if f0method != "rmvpe_gpu":
- cmd = (
- '"%s" infer/modules/train/extract/extract_f0_print.py "%s/logs/%s" %s %s'
- % (
- config.python_cmd,
- now_dir,
- exp_dir,
- n_p,
- f0method,
- echl,
- )
- )
- logger.info(cmd)
- p = Popen(
- cmd, shell=True, cwd=now_dir
- ) # , stdin=PIPE, stdout=PIPE,stderr=PIPE
- ###煞笔gr, popen read都非得全跑完了再一次性读取, 不用gr就正常读一句输出一句;只能额外弄出一个文本流定时读
- done = [False]
- threading.Thread(
- target=if_done,
- args=(
- done,
- p,
- ),
- ).start()
- else:
- if gpus_rmvpe != "-":
- gpus_rmvpe = gpus_rmvpe.split("-")
- leng = len(gpus_rmvpe)
- ps = []
- for idx, n_g in enumerate(gpus_rmvpe):
- cmd = (
- '"%s" infer/modules/train/extract/extract_f0_rmvpe.py %s %s %s "%s/logs/%s" %s '
- % (
- config.python_cmd,
- leng,
- idx,
- n_g,
- now_dir,
- exp_dir,
- config.is_half,
- )
- )
- logger.info(cmd)
- p = Popen(
- cmd, shell=True, cwd=now_dir
- ) # , shell=True, stdin=PIPE, stdout=PIPE, stderr=PIPE, cwd=now_dir
- ps.append(p)
- ###煞笔gr, popen read都非得全跑完了再一次性读取, 不用gr就正常读一句输出一句;只能额外弄出一个文本流定时读
- done = [False]
- threading.Thread(
- target=if_done_multi, #
- args=(
- done,
- ps,
- ),
- ).start()
- else:
- cmd = (
- config.python_cmd
- + ' infer/modules/train/extract/extract_f0_rmvpe_dml.py "%s/logs/%s" '
- % (
- now_dir,
- exp_dir,
- )
- )
- logger.info(cmd)
- p = Popen(
- cmd, shell=True, cwd=now_dir
- ) # , shell=True, stdin=PIPE, stdout=PIPE, stderr=PIPE, cwd=now_dir
- p.wait()
- done = [True]
- while 1:
- with open(
- "%s/logs/%s/extract_f0_feature.log" % (now_dir, exp_dir), "r"
- ) as f:
- yield (f.read())
- sleep(1)
- if done[0]:
- break
- with open("%s/logs/%s/extract_f0_feature.log" % (now_dir, exp_dir), "r") as f:
- log = f.read()
- logger.info(log)
- yield log
- ####对不同part分别开多进程
- """
- n_part=int(sys.argv[1])
- i_part=int(sys.argv[2])
- i_gpu=sys.argv[3]
- exp_dir=sys.argv[4]
- os.environ["CUDA_VISIBLE_DEVICES"]=str(i_gpu)
- """
- leng = len(gpus)
- ps = []
- for idx, n_g in enumerate(gpus):
- cmd = (
- '"%s" infer/modules/train/extract_feature_print.py %s %s %s %s "%s/logs/%s" %s'
- % (
- config.python_cmd,
- config.device,
- leng,
- idx,
- n_g,
- now_dir,
- exp_dir,
- version19,
- )
+with gr.Blocks(theme=my_applio, title="Applio") as Applio:
+ gr.Markdown("# Applio")
+ gr.Markdown("### From the first Applio to the last")
+ gr.Markdown(
+ i18n(
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience."
)
- logger.info(cmd)
- p = Popen(
- cmd, shell=True, cwd=now_dir
- ) # , shell=True, stdin=PIPE, stdout=PIPE, stderr=PIPE, cwd=now_dir
- ps.append(p)
- ###煞笔gr, popen read都非得全跑完了再一次性读取, 不用gr就正常读一句输出一句;只能额外弄出一个文本流定时读
- done = [False]
- threading.Thread(
- target=if_done_multi,
- args=(
- done,
- ps,
- ),
- ).start()
- while 1:
- with open("%s/logs/%s/extract_f0_feature.log" % (now_dir, exp_dir), "r") as f:
- yield (f.read())
- sleep(1)
- if done[0]:
- break
- with open("%s/logs/%s/extract_f0_feature.log" % (now_dir, exp_dir), "r") as f:
- log = f.read()
- logger.info(log)
- yield log
-
-def get_pretrained_models(path_str, f0_str, sr2):
- if_pretrained_generator_exist = os.access(
- "assets/pretrained%s/%sG%s.pth" % (path_str, f0_str, sr2), os.F_OK
- )
- if_pretrained_discriminator_exist = os.access(
- "assets/pretrained%s/%sD%s.pth" % (path_str, f0_str, sr2), os.F_OK
- )
- if not if_pretrained_generator_exist:
- logger.warn(
- "assets/pretrained%s/%sG%s.pth not exist, will not use pretrained model",
- path_str,
- f0_str,
- sr2,
- )
- if not if_pretrained_discriminator_exist:
- logger.warn(
- "assets/pretrained%s/%sD%s.pth not exist, will not use pretrained model",
- path_str,
- f0_str,
- sr2,
- )
- return (
- "assets/pretrained%s/%sG%s.pth" % (path_str, f0_str, sr2)
- if if_pretrained_generator_exist
- else "",
- "assets/pretrained%s/%sD%s.pth" % (path_str, f0_str, sr2)
- if if_pretrained_discriminator_exist
- else "",
- )
-
-def change_sr2(sr2, if_f0_3, version19):
- path_str = "" if version19 == "v1" else "_v2"
- f0_str = "f0" if if_f0_3 else ""
- return get_pretrained_models(path_str, f0_str, sr2)
-
-
-def change_version19(sr2, if_f0_3, version19):
- path_str = "" if version19 == "v1" else "_v2"
- if sr2 == "32k" and version19 == "v1":
- sr2 = "40k"
- to_return_sr2 = (
- {"choices": ["40k", "48k"], "__type__": "update", "value": sr2}
- if version19 == "v1"
- else {"choices": ["40k", "48k", "32k"], "__type__": "update", "value": sr2}
- )
- f0_str = "f0" if if_f0_3 else ""
- return (
- *get_pretrained_models(path_str, f0_str, sr2),
- to_return_sr2,
- )
-
-
-def change_f0(if_f0_3, sr2, version19): # f0method8,pretrained_G14,pretrained_D15
- path_str = "" if version19 == "v1" else "_v2"
- return (
- {"visible": if_f0_3, "__type__": "update"},
- *get_pretrained_models(path_str, "f0", sr2),
- )
-
-
-global log_interval
-
-def set_log_interval(exp_dir, batch_size12):
- log_interval = 1
- folder_path = os.path.join(exp_dir, "1_16k_wavs")
-
- if os.path.isdir(folder_path):
- wav_files_num = len(glob1(folder_path,"*.wav"))
-
- if wav_files_num > 0:
- log_interval = math.ceil(wav_files_num / batch_size12)
- if log_interval > 1:
- log_interval += 1
-
- return log_interval
-
-global PID, PROCESS
-
-def click_train(
- exp_dir1,
- sr2,
- if_f0_3,
- spk_id5,
- save_epoch10,
- total_epoch11,
- batch_size12,
- if_save_latest13,
- pretrained_G14,
- pretrained_D15,
- gpus16,
- if_cache_gpu17,
- if_save_every_weights18,
- version19,
-):
- CSVutil("csvdb/stop.csv", "w+", "formanting", False)
- # 生成filelist
- exp_dir = "%s/logs/%s" % (now_dir, exp_dir1)
- os.makedirs(exp_dir, exist_ok=True)
- gt_wavs_dir = "%s/0_gt_wavs" % (exp_dir)
- feature_dir = (
- "%s/3_feature256" % (exp_dir)
- if version19 == "v1"
- else "%s/3_feature768" % (exp_dir)
)
- if if_f0_3:
- f0_dir = "%s/2a_f0" % (exp_dir)
- f0nsf_dir = "%s/2b-f0nsf" % (exp_dir)
- names = (
- set([name.split(".")[0] for name in os.listdir(gt_wavs_dir)])
- & set([name.split(".")[0] for name in os.listdir(feature_dir)])
- & set([name.split(".")[0] for name in os.listdir(f0_dir)])
- & set([name.split(".")[0] for name in os.listdir(f0nsf_dir)])
- )
- else:
- names = set([name.split(".")[0] for name in os.listdir(gt_wavs_dir)]) & set(
- [name.split(".")[0] for name in os.listdir(feature_dir)]
- )
- opt = []
- for name in names:
- if if_f0_3:
- opt.append(
- "%s/%s.wav|%s/%s.npy|%s/%s.wav.npy|%s/%s.wav.npy|%s"
- % (
- gt_wavs_dir.replace("\\", "\\\\"),
- name,
- feature_dir.replace("\\", "\\\\"),
- name,
- f0_dir.replace("\\", "\\\\"),
- name,
- f0nsf_dir.replace("\\", "\\\\"),
- name,
- spk_id5,
- )
- )
- else:
- opt.append(
- "%s/%s.wav|%s/%s.npy|%s"
- % (
- gt_wavs_dir.replace("\\", "\\\\"),
- name,
- feature_dir.replace("\\", "\\\\"),
- name,
- spk_id5,
- )
- )
- fea_dim = 256 if version19 == "v1" else 768
- if if_f0_3:
- for _ in range(2):
- opt.append(
- "%s/logs/mute/0_gt_wavs/mute%s.wav|%s/logs/mute/3_feature%s/mute.npy|%s/logs/mute/2a_f0/mute.wav.npy|%s/logs/mute/2b-f0nsf/mute.wav.npy|%s"
- % (now_dir, sr2, now_dir, fea_dim, now_dir, now_dir, spk_id5)
- )
- else:
- for _ in range(2):
- opt.append(
- "%s/logs/mute/0_gt_wavs/mute%s.wav|%s/logs/mute/3_feature%s/mute.npy|%s"
- % (now_dir, sr2, now_dir, fea_dim, spk_id5)
- )
- shuffle(opt)
- with open("%s/filelist.txt" % exp_dir, "w") as f:
- f.write("\n".join(opt))
- logger.debug("Write filelist done")
- # 生成config#无需生成config
- # cmd = python_cmd + " train_nsf_sim_cache_sid_load_pretrain.py -e mi-test -sr 40k -f0 1 -bs 4 -g 0 -te 10 -se 5 -pg pretrained/f0G40k.pth -pd pretrained/f0D40k.pth -l 1 -c 0"
- logger.info("Use gpus: %s", str(gpus16))
- if pretrained_G14 == "":
- logger.info("No pretrained Generator")
- if pretrained_D15 == "":
- logger.info("No pretrained Discriminator")
- if version19 == "v1" or sr2 == "40k":
- config_path = "v1/%s.json" % sr2
- else:
- config_path = "v2/%s.json" % sr2
- config_save_path = os.path.join(exp_dir, "config.json")
- if not pathlib.Path(config_save_path).exists():
- with open(config_save_path, "w", encoding="utf-8") as f:
- json.dump(
- config.json_config[config_path],
- f,
- ensure_ascii=False,
- indent=4,
- sort_keys=True,
- )
- f.write("\n")
- if gpus16:
- cmd = (
- '"%s" infer/modules/train/train.py -e "%s" -sr %s -f0 %s -bs %s -g %s -te %s -se %s %s %s -l %s -c %s -sw %s -v %s'
- % (
- config.python_cmd,
- exp_dir1,
- sr2,
- 1 if if_f0_3 else 0,
- batch_size12,
- gpus16,
- total_epoch11,
- save_epoch10,
- "-pg %s" % pretrained_G14 if pretrained_G14 != "" else "",
- "-pd %s" % pretrained_D15 if pretrained_D15 != "" else "",
- 1 if if_save_latest13 == True else 0,
- 1 if if_cache_gpu17 == True else 0,
- 1 if if_save_every_weights18 == True else 0,
- version19,
- )
+ gr.Markdown(
+ i18n(
+ "[Support](https://discord.gg/IAHispano) — [Discord Bot](https://discord.com/oauth2/authorize?client_id=1144714449563955302&permissions=1376674695271&scope=bot%20applications.commands) — [Find Voices](https://applio.org/models) — [GitHub](https://github.com/IAHispano/Applio)"
)
- else:
- cmd = (
- '"%s" infer/modules/train/train.py -e "%s" -sr %s -f0 %s -bs %s -te %s -se %s %s %s -l %s -c %s -sw %s -v %s'
- % (
- config.python_cmd,
- exp_dir1,
- sr2,
- 1 if if_f0_3 else 0,
- batch_size12,
- total_epoch11,
- save_epoch10,
- "-pg %s" % pretrained_G14 if pretrained_G14 != "" else "",
- "-pd %s" % pretrained_D15 if pretrained_D15 != "" else "",
- 1 if if_save_latest13 == True else 0,
- 1 if if_cache_gpu17 == True else 0,
- 1 if if_save_every_weights18 == True else 0,
- version19,
- )
- )
- logger.info(cmd)
- global p
- p = Popen(cmd, shell=True, cwd=now_dir)
- global PID
- PID = p.pid
-
- p.wait()
-
- return i18n("Training is done, check train.log"), {"visible": False, "__type__": "update"}, {"visible": True, "__type__": "update"}
-
-
-def train_index(exp_dir1, version19):
- # exp_dir = "%s/logs/%s" % (now_dir, exp_dir1)
- exp_dir = "logs/%s" % (exp_dir1)
- os.makedirs(exp_dir, exist_ok=True)
- feature_dir = (
- "%s/3_feature256" % (exp_dir)
- if version19 == "v1"
- else "%s/3_feature768" % (exp_dir)
)
- if not os.path.exists(feature_dir):
- return "请先进行特征提取!"
- listdir_res = list(os.listdir(feature_dir))
- if len(listdir_res) == 0:
- return "请先进行特征提取！"
- infos = []
- npys = []
- for name in sorted(listdir_res):
- phone = np.load("%s/%s" % (feature_dir, name))
- npys.append(phone)
- big_npy = np.concatenate(npys, 0)
- big_npy_idx = np.arange(big_npy.shape[0])
- np.random.shuffle(big_npy_idx)
- big_npy = big_npy[big_npy_idx]
- if big_npy.shape[0] > 2e5:
- infos.append("Trying doing kmeans %s shape to 10k centers." % big_npy.shape[0])
- yield "\n".join(infos)
- try:
- big_npy = (
- MiniBatchKMeans(
- n_clusters=10000,
- verbose=True,
- batch_size=256 * config.n_cpu,
- compute_labels=False,
- init="random",
- )
- .fit(big_npy)
- .cluster_centers_
- )
- except:
- info = traceback.format_exc()
- logger.info(info)
- infos.append(info)
- yield "\n".join(infos)
+ with gr.Tab(i18n("Inference")):
+ inference_tab()
- np.save("%s/total_fea.npy" % exp_dir, big_npy)
- n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])), big_npy.shape[0] // 39)
- infos.append("%s,%s" % (big_npy.shape, n_ivf))
- yield "\n".join(infos)
- index = faiss.index_factory(256 if version19 == "v1" else 768, "IVF%s,Flat" % n_ivf)
- # index = faiss.index_factory(256if version19=="v1"else 768, "IVF%s,PQ128x4fs,RFlat"%n_ivf)
- infos.append("training")
- yield "\n".join(infos)
- index_ivf = faiss.extract_index_ivf(index) #
- index_ivf.nprobe = 1
- index.train(big_npy)
- faiss.write_index(
- index,
- "%s/trained_IVF%s_Flat_nprobe_%s_%s_%s.index"
- % (exp_dir, n_ivf, index_ivf.nprobe, exp_dir1, version19),
- )
-
- infos.append("adding")
- yield "\n".join(infos)
- batch_size_add = 8192
- for i in range(0, big_npy.shape[0], batch_size_add):
- index.add(big_npy[i : i + batch_size_add])
- faiss.write_index(
- index,
- "%s/added_IVF%s_Flat_nprobe_%s_%s_%s.index"
- % (exp_dir, n_ivf, index_ivf.nprobe, exp_dir1, version19),
- )
- infos.append(
- "Successful Index Construction，added_IVF%s_Flat_nprobe_%s_%s_%s.index"
- % (n_ivf, index_ivf.nprobe, exp_dir1, version19)
- )
- # faiss.write_index(index, '%s/added_IVF%s_Flat_FastScan_%s.index'%(exp_dir,n_ivf,version19))
- # infos.append("成功构建索引，added_IVF%s_Flat_FastScan_%s.index"%(n_ivf,version19))
- yield "\n".join(infos)
+ with gr.Tab(i18n("TTS")):
+ tts_tab()
-def change_info_(ckpt_path):
- if not os.path.exists(ckpt_path.replace(os.path.basename(ckpt_path), "train.log")):
- return {"__type__": "update"}, {"__type__": "update"}, {"__type__": "update"}
- try:
- with open(
- ckpt_path.replace(os.path.basename(ckpt_path), "train.log"), "r"
- ) as f:
- info = eval(f.read().strip("\n").split("\n")[0].split("\t")[-1])
- sr, f0 = info["sample_rate"], info["if_f0"]
- version = "v2" if ("version" in info and info["version"] == "v2") else "v1"
- return sr, str(f0), version
- except:
- traceback.print_exc()
- return {"__type__": "update"}, {"__type__": "update"}, {"__type__": "update"}
+ with gr.Tab(i18n("Download")):
+ download_tab()
-F0GPUVisible = config.dml == False
+def launch_gradio():
+ Applio.launch()
-def change_f0_method(f0method8):
- if f0method8 == "rmvpe_gpu":
- visible = F0GPUVisible
- else:
- visible = False
- return {"visible": visible, "__type__": "update"}
-
-
-
-def export_onnx(model_path, exported_path):
- device = torch.device("cpu")
- checkpoint = torch.load(model_path, map_location=device)
- vec_channels = 256 if checkpoint.get("version", "v1") == "v1" else 768
-
- test_inputs = {
- "phone": torch.rand(1, 200, vec_channels),
- "phone_lengths": torch.LongTensor([200]),
- "pitch": torch.randint(5, 255, (1, 200)),
- "pitchf": torch.rand(1, 200),
- "ds": torch.zeros(1).long(),
- "rnd": torch.rand(1, 192, 200)
- }
-
- checkpoint["config"][-3] = checkpoint["weight"]["emb_g.weight"].shape[0]
- net_g = SynthesizerTrnMsNSFsidM(*checkpoint["config"], is_half=False, version=checkpoint.get("version", "v1"))
-
- net_g.load_state_dict(checkpoint["weight"], strict=False)
- net_g = net_g.to(device)
-
- dynamic_axes = {"phone": [1], "pitch": [1], "pitchf": [1], "rnd": [2]}
-
- torch.onnx.export(
- net_g,
- tuple(value.to(device) for value in test_inputs.values()),
- exported_path,
- dynamic_axes=dynamic_axes,
- do_constant_folding=False,
- opset_version=13,
- verbose=False,
- input_names=list(test_inputs.keys()),
- output_names=["audio"],
- )
- return "Finished"
-
-
-
-import re as regex
-import scipy.io.wavfile as wavfile
-
-cli_current_page = "HOME"
-
-
-def cli_split_command(com):
- exp = r'(?:(?<=\s)|^)"(.*?)"(?=\s|$)|(\S+)'
- split_array = regex.findall(exp, com)
- split_array = [group[0] if group[0] else group[1] for group in split_array]
- return split_array
-
-
-def execute_generator_function(genObject):
- for _ in genObject:
- pass
-
-
-def cli_infer(com):
- # get VC first
- com = cli_split_command(com)
- model_name = com[0]
- source_audio_path = com[1]
- output_file_name = com[2]
- feature_index_path = com[3]
- f0_file = None # Not Implemented Yet
-
- # Get parameters for inference
- speaker_id = int(com[4])
- transposition = float(com[5])
- f0_method = com[6]
- crepe_hop_length = int(com[7])
- harvest_median_filter = int(com[8])
- resample = int(com[9])
- mix = float(com[10])
- feature_ratio = float(com[11])
- protection_amnt = float(com[12])
- protect1 = 0.5
-
- if com[14] == "False" or com[14] == "false":
- DoFormant = False
- Quefrency = 0.0
- Timbre = 0.0
- CSVutil(
- "csvdb/formanting.csv", "w+", "formanting", DoFormant, Quefrency, Timbre
- )
-
- else:
- DoFormant = True
- Quefrency = float(com[15])
- Timbre = float(com[16])
- CSVutil(
- "csvdb/formanting.csv", "w+", "formanting", DoFormant, Quefrency, Timbre
- )
-
- print("Mangio-RVC-Fork Infer-CLI: Starting the inference...")
- vc_data = vc.get_vc(model_name, protection_amnt, protect1)
- print(vc_data)
- print("Mangio-RVC-Fork Infer-CLI: Performing inference...")
- conversion_data = vc.vc_single(
- speaker_id,
- source_audio_path,
- source_audio_path,
- transposition,
- f0_file,
- f0_method,
- feature_index_path,
- feature_index_path,
- feature_ratio,
- harvest_median_filter,
- resample,
- mix,
- protection_amnt,
- crepe_hop_length,
- )
- if "Success." in conversion_data[0]:
- print(
- "Mangio-RVC-Fork Infer-CLI: Inference succeeded. Writing to %s/%s..."
- % ("audio-outputs", output_file_name)
- )
- wavfile.write(
- "%s/%s" % ("audio-outputs", output_file_name),
- conversion_data[1][0],
- conversion_data[1][1],
- )
- print(
- "Mangio-RVC-Fork Infer-CLI: Finished! Saved output to %s/%s"
- % ("audio-outputs", output_file_name)
- )
- else:
- print("Mangio-RVC-Fork Infer-CLI: Inference failed. Here's the traceback: ")
- print(conversion_data[0])
-
-
-def cli_pre_process(com):
- com = cli_split_command(com)
- model_name = com[0]
- trainset_directory = com[1]
- sample_rate = com[2]
- num_processes = int(com[3])
-
- print("Mangio-RVC-Fork Pre-process: Starting...")
- generator = preprocess_dataset(
- trainset_directory, model_name, sample_rate, num_processes
- )
- execute_generator_function(generator)
- print("Mangio-RVC-Fork Pre-process: Finished")
-
-
-def cli_extract_feature(com):
- com = cli_split_command(com)
- model_name = com[0]
- gpus = com[1]
- num_processes = int(com[2])
- has_pitch_guidance = True if (int(com[3]) == 1) else False
- f0_method = com[4]
- crepe_hop_length = int(com[5])
- version = com[6] # v1 or v2
-
- print("Mangio-RVC-CLI: Extract Feature Has Pitch: " + str(has_pitch_guidance))
- print("Mangio-RVC-CLI: Extract Feature Version: " + str(version))
- print("Mangio-RVC-Fork Feature Extraction: Starting...")
- generator = extract_f0_feature(
- gpus,
- num_processes,
- f0_method,
- has_pitch_guidance,
- model_name,
- version,
- crepe_hop_length,
- )
- execute_generator_function(generator)
- print("Mangio-RVC-Fork Feature Extraction: Finished")
-
-
-def cli_train(com):
- com = cli_split_command(com)
- model_name = com[0]
- sample_rate = com[1]
- has_pitch_guidance = True if (int(com[2]) == 1) else False
- speaker_id = int(com[3])
- save_epoch_iteration = int(com[4])
- total_epoch = int(com[5]) # 10000
- batch_size = int(com[6])
- gpu_card_slot_numbers = com[7]
- if_save_latest = True if (int(com[8]) == 1) else False
- if_cache_gpu = True if (int(com[9]) == 1) else False
- if_save_every_weight = True if (int(com[10]) == 1) else False
- version = com[11]
-
- pretrained_base = "pretrained/" if version == "v1" else "pretrained_v2/"
-
- g_pretrained_path = "%sf0G%s.pth" % (pretrained_base, sample_rate)
- d_pretrained_path = "%sf0D%s.pth" % (pretrained_base, sample_rate)
-
- print("Mangio-RVC-Fork Train-CLI: Training...")
- click_train(
- model_name,
- sample_rate,
- has_pitch_guidance,
- speaker_id,
- save_epoch_iteration,
- total_epoch,
- batch_size,
- if_save_latest,
- g_pretrained_path,
- d_pretrained_path,
- gpu_card_slot_numbers,
- if_cache_gpu,
- if_save_every_weight,
- version,
- )
-
-
-def cli_train_feature(com):
- com = cli_split_command(com)
- model_name = com[0]
- version = com[1]
- print("Mangio-RVC-Fork Train Feature Index-CLI: Training... Please wait")
- generator = train_index(model_name, version)
- execute_generator_function(generator)
- print("Mangio-RVC-Fork Train Feature Index-CLI: Done!")
-
-
-def cli_extract_model(com):
- com = cli_split_command(com)
- model_path = com[0]
- save_name = com[1]
- sample_rate = com[2]
- has_pitch_guidance = com[3]
- info = com[4]
- version = com[5]
- extract_small_model_process = extract_small_model(
- model_path, save_name, sample_rate, has_pitch_guidance, info, version
- )
- if extract_small_model_process == "Success.":
- print("Mangio-RVC-Fork Extract Small Model: Success!")
- else:
- print(str(extract_small_model_process))
- print("Mangio-RVC-Fork Extract Small Model: Failed!")
-
-
-def preset_apply(preset, qfer, tmbr):
- if str(preset) != "":
- with open(str(preset), "r") as p:
- content = p.readlines()
- qfer, tmbr = content[0].split("\n")[0], content[1]
- formant_apply(qfer, tmbr)
- else:
- pass
- return (
- {"value": qfer, "__type__": "update"},
- {"value": tmbr, "__type__": "update"},
- )
-
-
-def print_page_details():
- if cli_current_page == "HOME":
- print(
- "\n go home : Takes you back to home with a navigation list."
- "\n go infer : Takes you to inference command execution."
- "\n go pre-process : Takes you to training step.1) pre-process command execution."
- "\n go extract-feature : Takes you to training step.2) extract-feature command execution."
- "\n go train : Takes you to training step.3) being or continue training command execution."
- "\n go train-feature : Takes you to the train feature index command execution."
- "\n go extract-model : Takes you to the extract small model command execution."
- )
- elif cli_current_page == "INFER":
- print(
- "\n arg 1) model name with .pth in ./weights: mi-test.pth"
- "\n arg 2) source audio path: myFolder\\MySource.wav"
- "\n arg 3) output file name to be placed in './audio-outputs': MyTest.wav"
- "\n arg 4) feature index file path: logs/mi-test/added_IVF3042_Flat_nprobe_1.index"
- "\n arg 5) speaker id: 0"
- "\n arg 6) transposition: 0"
- "\n arg 7) f0 method: harvest (pm, harvest, crepe, crepe-tiny, hybrid[x,x,x,x], mangio-crepe, mangio-crepe-tiny, rmvpe)"
- "\n arg 8) crepe hop length: 160"
- "\n arg 9) harvest median filter radius: 3 (0-7)"
- "\n arg 10) post resample rate: 0"
- "\n arg 11) mix volume envelope: 1"
- "\n arg 12) feature index ratio: 0.78 (0-1)"
- "\n arg 13) Voiceless Consonant Protection (Less Artifact): 0.33 (Smaller number = more protection. 0.50 means Dont Use.)"
- "\n arg 14) Whether to formant shift the inference audio before conversion: False (if set to false, you can ignore setting the quefrency and timbre values for formanting)"
- "\n arg 15)* Quefrency for formanting: 8.0 (no need to set if arg14 is False/false)"
- "\n arg 16)* Timbre for formanting: 1.2 (no need to set if arg14 is False/false) \n"
- "\nExample: mi-test.pth saudio/Sidney.wav myTest.wav logs/mi-test/added_index.index 0 -2 harvest 160 3 0 1 0.95 0.33 0.45 True 8.0 1.2"
- )
- elif cli_current_page == "PRE-PROCESS":
- print(
- "\n arg 1) Model folder name in ./logs: mi-test"
- "\n arg 2) Trainset directory: mydataset (or) E:\\my-data-set"
- "\n arg 3) Sample rate: 40k (32k, 40k, 48k)"
- "\n arg 4) Number of CPU threads to use: 8 \n"
- "\nExample: mi-test mydataset 40k 24"
- )
- elif cli_current_page == "EXTRACT-FEATURE":
- print(
- "\n arg 1) Model folder name in ./logs: mi-test"
- "\n arg 2) Gpu card slot: 0 (0-1-2 if using 3 GPUs)"
- "\n arg 3) Number of CPU threads to use: 8"
- "\n arg 4) Has Pitch Guidance?: 1 (0 for no, 1 for yes)"
- "\n arg 5) f0 Method: harvest (pm, harvest, dio, crepe)"
- "\n arg 6) Crepe hop length: 128"
- "\n arg 7) Version for pre-trained models: v2 (use either v1 or v2)\n"
- "\nExample: mi-test 0 24 1 harvest 128 v2"
- )
- elif cli_current_page == "TRAIN":
- print(
- "\n arg 1) Model folder name in ./logs: mi-test"
- "\n arg 2) Sample rate: 40k (32k, 40k, 48k)"
- "\n arg 3) Has Pitch Guidance?: 1 (0 for no, 1 for yes)"
- "\n arg 4) speaker id: 0"
- "\n arg 5) Save epoch iteration: 50"
- "\n arg 6) Total epochs: 10000"
- "\n arg 7) Batch size: 8"
- "\n arg 8) Gpu card slot: 0 (0-1-2 if using 3 GPUs)"
- "\n arg 9) Save only the latest checkpoint: 0 (0 for no, 1 for yes)"
- "\n arg 10) Whether to cache training set to vram: 0 (0 for no, 1 for yes)"
- "\n arg 11) Save extracted small model every generation?: 0 (0 for no, 1 for yes)"
- "\n arg 12) Model architecture version: v2 (use either v1 or v2)\n"
- "\nExample: mi-test 40k 1 0 50 10000 8 0 0 0 0 v2"
- )
- elif cli_current_page == "TRAIN-FEATURE":
- print(
- "\n arg 1) Model folder name in ./logs: mi-test"
- "\n arg 2) Model architecture version: v2 (use either v1 or v2)\n"
- "\nExample: mi-test v2"
- )
- elif cli_current_page == "EXTRACT-MODEL":
- print(
- "\n arg 1) Model Path: logs/mi-test/G_168000.pth"
- "\n arg 2) Model save name: MyModel"
- "\n arg 3) Sample rate: 40k (32k, 40k, 48k)"
- "\n arg 4) Has Pitch Guidance?: 1 (0 for no, 1 for yes)"
- '\n arg 5) Model information: "My Model"'
- "\n arg 6) Model architecture version: v2 (use either v1 or v2)\n"
- '\nExample: logs/mi-test/G_168000.pth MyModel 40k 1 "Created by Cole Mangio" v2'
- )
-
-def change_page(page):
- global cli_current_page
- cli_current_page = page
- return 0
-
-def execute_command(com):
- if com == "go home":
- return change_page("HOME")
- elif com == "go infer":
- return change_page("INFER")
- elif com == "go pre-process":
- return change_page("PRE-PROCESS")
- elif com == "go extract-feature":
- return change_page("EXTRACT-FEATURE")
- elif com == "go train":
- return change_page("TRAIN")
- elif com == "go train-feature":
- return change_page("TRAIN-FEATURE")
- elif com == "go extract-model":
- return change_page("EXTRACT-MODEL")
- else:
- if com[:3] == "go ":
- print("page '%s' does not exist!" % com[3:])
- return 0
-
- if cli_current_page == "INFER":
- cli_infer(com)
- elif cli_current_page == "PRE-PROCESS":
- cli_pre_process(com)
- elif cli_current_page == "EXTRACT-FEATURE":
- cli_extract_feature(com)
- elif cli_current_page == "TRAIN":
- cli_train(com)
- elif cli_current_page == "TRAIN-FEATURE":
- cli_train_feature(com)
- elif cli_current_page == "EXTRACT-MODEL":
- cli_extract_model(com)
-
-def cli_navigation_loop():
- while True:
- print("\nYou are currently in '%s':" % cli_current_page)
- print_page_details()
- command = input("%s: " % cli_current_page)
- try:
- execute_command(command)
- except:
- print(traceback.format_exc())
-
-
-if config.is_cli:
- print("\n\nMangio-RVC-Fork v2 CLI App!\n")
- print(
- "Welcome to the CLI version of RVC. Please read the documentation on https://github.com/Mangio621/Mangio-RVC-Fork (README.MD) to understand how to use this app.\n"
- )
- cli_navigation_loop()
-
-
-
-
-
-def switch_pitch_controls(f0method0):
- is_visible = f0method0 != 'rmvpe'
-
- if rvc_globals.NotesOrHertz:
- return (
- {"visible": False, "__type__": "update"},
- {"visible": is_visible, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": is_visible, "__type__": "update"}
- )
- else:
- return (
- {"visible": is_visible, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": is_visible, "__type__": "update"},
- {"visible": False, "__type__": "update"}
- )
-
-def match_index(sid0):
- picked = False
- # folder = sid0.split('.')[0]
-
- # folder = re.split(r'. |_', sid0)[0]
- folder = sid0.split(".")[0].split("_")[0]
- # folder_test = sid0.split('.')[0].split('_')[0].split('-')[0]
- parent_dir = "./logs/" + folder
- # print(parent_dir)
- if os.path.exists(parent_dir):
- # print('path exists')
- for filename in os.listdir(parent_dir.replace("\\", "/")):
- if filename.endswith(".index"):
- for i in range(len(indexes_list)):
- if indexes_list[i] == (
- os.path.join(("./logs/" + folder), filename).replace("\\", "/")
- ):
- # print('regular index found')
- break
- else:
- if indexes_list[i] == (
- os.path.join(
- ("./logs/" + folder.lower()), filename
- ).replace("\\", "/")
- ):
- # print('lowered index found')
- parent_dir = "./logs/" + folder.lower()
- break
- # elif (indexes_list[i]).casefold() == ((os.path.join(("./logs/" + folder), filename).replace('\\','/')).casefold()):
- # print('8')
- # parent_dir = "./logs/" + folder.casefold()
- # break
- # elif (indexes_list[i]) == ((os.path.join(("./logs/" + folder_test), filename).replace('\\','/'))):
- # parent_dir = "./logs/" + folder_test
- # print(parent_dir)
- # break
- # elif (indexes_list[i]) == (os.path.join(("./logs/" + folder_test.lower()), filename).replace('\\','/')):
- # parent_dir = "./logs/" + folder_test
- # print(parent_dir)
- # break
- # else:
- # #print('couldnt find index')
- # continue
-
- # print('all done')
- index_path = os.path.join(
- parent_dir.replace("\\", "/"), filename.replace("\\", "/")
- ).replace("\\", "/")
- # print(index_path)
- return (index_path, index_path)
-
- else:
- # print('nothing found')
- return ("", "")
-
-def stoptraining(mim):
- if int(mim) == 1:
- CSVutil("csvdb/stop.csv", "w+", "stop", "True")
- # p.terminate()
- # p.kill()
- try:
- os.kill(PID, signal.SIGTERM)
- except Exception as e:
- print(f"Couldn't click due to {e}")
- pass
- else:
- pass
-
- return (
- {"visible": False, "__type__": "update"},
- {"visible": True, "__type__": "update"},
- )
-
-weights_dir = 'weights/'
-
-def note_to_hz(note_name):
- SEMITONES = {'C': -9, 'C#': -8, 'D': -7, 'D#': -6, 'E': -5, 'F': -4, 'F#': -3, 'G': -2, 'G#': -1, 'A': 0, 'A#': 1, 'B': 2}
- pitch_class, octave = note_name[:-1], int(note_name[-1])
- semitone = SEMITONES[pitch_class]
- note_number = 12 * (octave - 4) + semitone
- frequency = 440.0 * (2.0 ** (1.0/12)) ** note_number
- return frequency
-
-def save_to_wav(record_button):
- if record_button is None:
- pass
- else:
- path_to_file=record_button
- new_name = datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S")+'.wav'
- new_path='./audios/'+new_name
- shutil.move(path_to_file,new_path)
- return new_name
-def save_to_wav2_edited(dropbox):
- if dropbox is None:
- pass
- else:
- file_path = dropbox.name
- target_path = os.path.join('audios', os.path.basename(file_path))
-
- if os.path.exists(target_path):
- os.remove(target_path)
- print('Replacing old dropdown file...')
-
- shutil.move(file_path, target_path)
- return
-def save_to_wav2(dropbox):
- file_path = dropbox.name
- target_path = os.path.join('audios', os.path.basename(file_path))
-
- if os.path.exists(target_path):
- os.remove(target_path)
- print('Replacing old dropdown file...')
-
- shutil.move(file_path, target_path)
- return target_path
-
-from gtts import gTTS
-import edge_tts
-import asyncio
-
-
-
-
-def custom_voice(
- _values, # filter indices
- audio_files, # all audio files
- model_voice_path='',
- transpose=0,
- f0method='pm',
- index_rate_=float(0.66),
- crepe_hop_length_=float(64),
- f0_autotune=False,
- file_index='',
- file_index2='',
- ):
-
- vc.get_vc(model_voice_path)
-
-
- for _value_item in _values:
- filename = "audio2/"+audio_files[_value_item] if _value_item != "converted_tts" else audio_files[0]
- #filename = "audio2/"+audio_files[_value_item]
- try:
- print(audio_files[_value_item], model_voice_path)
- except:
- pass
- info_, (sample_, audio_output_) = vc.vc_single_dont_save(
- sid=0,
- input_audio_path0=filename, #f"audio2/{filename}",
- input_audio_path1=filename, #f"audio2/{filename}",
- f0_up_key=transpose, # transpose for m to f and reverse 0 12
- f0_file=None,
- f0_method= f0method,
- file_index= file_index, # dir pwd?
- file_index2= file_index2,
- # file_big_npy1,
- index_rate= index_rate_,
- filter_radius= int(3),
- resample_sr= int(0),
- rms_mix_rate= float(0.25),
- protect= float(0.33),
- crepe_hop_length= crepe_hop_length_,
- f0_autotune=f0_autotune,
- f0_min=50,
- note_min=50,
- f0_max=1100,
- note_max=1100
- )
-
- sf.write(
- file= filename, #f"audio2/{filename}",
- samplerate=sample_,
- data=audio_output_
- )
-def cast_to_device(tensor, device):
- try:
- return tensor.to(device)
- except Exception as e:
- print(e)
- return tensor
-
-
-def __bark__(text, voice_preset):
- os.makedirs(os.path.join(now_dir,"tts"), exist_ok=True)
- from transformers import AutoProcessor, BarkModel
- device = "cuda:0" if torch.cuda.is_available() else "cpu"
- dtype = torch.float32 if "cpu" in device else torch.float16
- bark_processor = AutoProcessor.from_pretrained(
- "suno/bark-small",
- cache_dir=os.path.join(now_dir,"tts","suno/bark"),
- torch_dtype=dtype)
- bark_model = BarkModel.from_pretrained(
- "suno/bark-small",
- cache_dir=os.path.join(now_dir,"tts","suno/bark"),
- torch_dtype=dtype).to(device)
- # bark_model.enable_cpu_offload()
- inputs = bark_processor(
- text=[text],
- return_tensors="pt",
- voice_preset=voice_preset
- )
- tensor_dict = {k: cast_to_device(v,device) if hasattr(v,"to") else v for k, v in inputs.items()}
- speech_values = bark_model.generate(**tensor_dict, do_sample=True)
- sampling_rate = bark_model.generation_config.sample_rate
- speech = speech_values.cpu().numpy().squeeze()
- return speech, sampling_rate
-
-
-
-def make_test(
- tts_text,
- tts_voice,
- model_path,
- index_path,
- transpose,
- f0_method,
- index_rate,
- crepe_hop_length,
- f0_autotune,
- tts_method
- ):
-
- if tts_voice == None:
- return
-
- filename = os.path.join(now_dir, "audio-outputs", "converted_tts.wav")
- if "SET_LIMIT" == os.getenv("DEMO"):
- if len(tts_text) > 60:
- tts_text = tts_text[:60]
- print("DEMO; limit to 60 characters")
-
- language = tts_voice[:2]
- if tts_method == "Edge-tts":
- try:
- #nest_asyncio.apply() # gradio;not
- asyncio.run(edge_tts.Communicate(tts_text, "-".join(tts_voice.split('-')[:-1])).save(filename))
- except:
- try:
- tts = gTTS(tts_text, lang=language)
- tts.save(filename)
- tts.save
- print(f'No audio was received. Please change the tts voice for {tts_voice}. USING gTTS.')
- except:
- tts = gTTS('a', lang=language)
- tts.save(filename)
- print('Error: Audio will be replaced.')
-
- os.system("cp audio-outputs/converted_tts.wav audio-outputs/real_tts.wav")
-
- custom_voice(
- ["converted_tts"], # filter indices
- ["audio-outputs/converted_tts.wav"], # all audio files
- model_voice_path=model_path,
- transpose=transpose,
- f0method=f0_method,
- index_rate_=index_rate,
- crepe_hop_length_=crepe_hop_length,
- f0_autotune=f0_autotune,
- file_index='',
- file_index2=index_path,
- )
- return os.path.join(now_dir, "audio-outputs", "converted_tts.wav"), os.path.join(now_dir, "audio-outputs", "real_tts.wav")
- elif tts_method == "Bark-tts":
- try:
-
- script = tts_text.replace("\n", " ").strip()
- sentences = sent_tokenize(script)
- print(sentences)
- silence = np.zeros(int(0.25 * SAMPLE_RATE))
- pieces = []
- nombre_archivo = os.path.join(now_dir, "audio-outputs", "bark_out.wav")
- for sentence in sentences:
- audio_array , _ = __bark__(sentence, tts_voice.split("-")[0])
- pieces += [audio_array, silence.copy()]
-
- sf.write(
- file= nombre_archivo,
- samplerate=SAMPLE_RATE,
- data=np.concatenate(pieces)
- )
- vc.get_vc(model_path)
- info_, (sample_, audio_output_) = vc.vc_single_dont_save(
- sid=0,
- input_audio_path0=os.path.join(now_dir, "audio-outputs", "bark_out.wav"), #f"audio2/{filename}",
- input_audio_path1=os.path.join(now_dir, "audio-outputs", "bark_out.wav"), #f"audio2/{filename}",
- f0_up_key=transpose, # transpose for m to f and reverse 0 12
- f0_file=None,
- f0_method=f0_method,
- file_index= '', # dir pwd?
- file_index2= index_path,
- # file_big_npy1,
- index_rate= index_rate,
- filter_radius= int(3),
- resample_sr= int(0),
- rms_mix_rate= float(0.25),
- protect= float(0.33),
- crepe_hop_length= crepe_hop_length,
- f0_autotune=f0_autotune,
- f0_min=50,
- note_min=50,
- f0_max=1100,
- note_max=1100
- )
- wavfile.write(os.path.join(now_dir, "audio-outputs", "converted_bark.wav"), rate=sample_, data=audio_output_)
- return os.path.join(now_dir, "audio-outputs", "converted_bark.wav"), nombre_archivo
-
- except Exception as e:
- print(f"{e}")
- return None, None
-
-
-
-
-
-
-def GradioSetup(UTheme=gr.themes.Soft()):
-
- default_weight = names[0] if names else ''
-
- with gr.Blocks(theme='JohnSmith9982/small_and_pretty', title="Applio") as app:
- gr.Markdown("🍏 Applio (Mangio-RVC-Fork HF)")
- gr.Markdown("More spaces: [Aesthetic_RVC_Inference_HF](https://huggingface.co/spaces/r3gm/Aesthetic_RVC_Inference_HF), [AICoverGen](https://huggingface.co/spaces/r3gm/AICoverGen), [Ultimate-Vocal-Remover-WebUI](https://huggingface.co/spaces/r3gm/Ultimate-Vocal-Remover-WebUI), [Advanced-RVC-Inference](https://huggingface.co/spaces/r3gm/Advanced-RVC-Inference)")
- gr.HTML("
The current space only uses CPU, so it's only for inference. If you have issues with the queue, I recommend duplicating the space.
")
- gr.Markdown(
- "[](https://huggingface.co/spaces/r3gm/RVC_HF?duplicate=true)\n\n"
- )
- with gr.Tabs():
- with gr.TabItem(i18n("Model Inference")):
- with gr.Row():
- sid0 = gr.Dropdown(label=i18n("Inferencing voice:"), choices=sorted(names), value=default_weight)
- refresh_button = gr.Button(i18n("Refresh"), variant="primary")
- clean_button = gr.Button(i18n("Unload voice to save GPU memory"), variant="primary")
- clean_button.click(fn=lambda: ({"value": "", "__type__": "update"}), inputs=[], outputs=[sid0])
-
-
- with gr.TabItem(i18n("Single")):
- with gr.Row():
- spk_item = gr.Slider(
- minimum=0,
- maximum=2333,
- step=1,
- label=i18n("Select Speaker/Singer ID:"),
- value=0,
- visible=False,
- interactive=True,
- )
-
-
- with gr.Group():
- with gr.Row():
- with gr.Column(): # First column for audio-related inputs
- dropbox = gr.File(label=i18n("Drag your audio here:"))
- record_button=gr.Audio(source="microphone", label=i18n("Or record an audio:"), type="filepath")
- input_audio0 = gr.Textbox(
- label=i18n("Manual path to the audio file to be processed"),
- value=os.path.join(now_dir, "audios", "someguy.mp3"),
- visible=False
- )
- input_audio1 = gr.Dropdown(
- label=i18n("Auto detect audio path and select from the dropdown:"),
- choices=sorted(audio_paths),
- value='',
- interactive=True,
- )
-
- input_audio1.select(fn=lambda:'',inputs=[],outputs=[input_audio0])
- input_audio0.input(fn=lambda:'',inputs=[],outputs=[input_audio1])
-
- dropbox.upload(fn=save_to_wav2, inputs=[dropbox], outputs=[input_audio0])
- dropbox.upload(fn=easy_infer.change_choices2, inputs=[], outputs=[input_audio1])
- record_button.change(fn=save_to_wav, inputs=[record_button], outputs=[input_audio0])
- record_button.change(fn=easy_infer.change_choices2, inputs=[], outputs=[input_audio1])
-
- best_match_index_path1 = match_index(sid0.value) # Get initial index from default sid0 (first voice model in list)
-
- with gr.Column(): # Second column for pitch shift and other options
- file_index2 = gr.Dropdown(
- label=i18n("Auto-detect index path and select from the dropdown:"),
- choices=get_indexes(),
- value=best_match_index_path1,
- interactive=True,
- allow_custom_value=True,
- )
- index_rate1 = gr.Slider(
- minimum=0,
- maximum=1,
- label=i18n("Search feature ratio:"),
- value=0.75,
- interactive=True,
- )
- refresh_button.click(
- fn=change_choices, inputs=[], outputs=[sid0, file_index2, input_audio1]
- )
- with gr.Column():
- vc_transform0 = gr.Number(
- label=i18n("Transpose (integer, number of semitones, raise by an octave: 12, lower by an octave: -12):"), value=0
- )
-
- # Create a checkbox for advanced settings
- advanced_settings_checkbox = gr.Checkbox(
- value=False,
- label=i18n("Advanced Settings"),
- interactive=True,
- )
-
- # Advanced settings container
- with gr.Column(visible=False) as advanced_settings: # Initially hidden
- with gr.Row(label = i18n("Advanced Settings"), open = False):
- with gr.Column():
- f0method0 = gr.Radio(
- label=i18n(
- "Select the pitch extraction algorithm:"
- ),
- choices=["pm", "harvest", "dio", "crepe", "crepe-tiny", "mangio-crepe", "mangio-crepe-tiny", "rmvpe", "rmvpe+"],
- value="rmvpe+",
- interactive=True,
- )
- f0_autotune = gr.Checkbox(
- label="Enable autotune",
- interactive=True
- )
- crepe_hop_length = gr.Slider(
- minimum=1,
- maximum=512,
- step=1,
- label=i18n("Mangio-Crepe Hop Length (Only applies to mangio-crepe): Hop length refers to the time it takes for the speaker to jump to a dramatic pitch. Lower hop lengths take more time to infer but are more pitch accurate."),
- value=120,
- interactive=True,
- visible=False,
- )
- filter_radius0 = gr.Slider(
- minimum=0,
- maximum=7,
- label=i18n("If >=3: apply median filtering to the harvested pitch results. The value represents the filter radius and can reduce breathiness."),
- value=3,
- step=1,
- interactive=True,
- )
-
- minpitch_slider = gr.Slider(
- label = i18n("Min pitch:"),
- info = i18n("Specify minimal pitch for inference [HZ]"),
- step = 0.1,
- minimum = 1,
- scale = 0,
- value = 50,
- maximum = 16000,
- interactive = True,
- visible = (not rvc_globals.NotesOrHertz) and (f0method0.value != 'rmvpe'),
- )
- minpitch_txtbox = gr.Textbox(
- label = i18n("Min pitch:"),
- info = i18n("Specify minimal pitch for inference [NOTE][OCTAVE]"),
- placeholder = "C5",
- visible = (rvc_globals.NotesOrHertz) and (f0method0.value != 'rmvpe'),
- interactive = True,
- )
-
- maxpitch_slider = gr.Slider(
- label = i18n("Max pitch:"),
- info = i18n("Specify max pitch for inference [HZ]"),
- step = 0.1,
- minimum = 1,
- scale = 0,
- value = 1100,
- maximum = 16000,
- interactive = True,
- visible = (not rvc_globals.NotesOrHertz) and (f0method0.value != 'rmvpe'),
- )
- maxpitch_txtbox = gr.Textbox(
- label = i18n("Max pitch:"),
- info = i18n("Specify max pitch for inference [NOTE][OCTAVE]"),
- placeholder = "C6",
- visible = (rvc_globals.NotesOrHertz) and (f0method0.value != 'rmvpe'),
- interactive = True,
- )
-
- with gr.Column():
- file_index1 = gr.Textbox(
- label=i18n("Feature search database file path:"),
- value="",
- interactive=True,
- )
-
- with gr.Accordion(label = i18n("Custom f0 [Root pitch] File"), open = False):
- f0_file = gr.File(label=i18n("F0 curve file (optional). One pitch per line. Replaces the default F0 and pitch modulation:"))
-
- f0method0.change(
- fn=lambda radio: (
- {
- "visible": radio in ['mangio-crepe', 'mangio-crepe-tiny'],
- "__type__": "update"
- }
- ),
- inputs=[f0method0],
- outputs=[crepe_hop_length]
- )
-
- f0method0.change(
- fn=switch_pitch_controls,
- inputs=[f0method0],
- outputs=[minpitch_slider, minpitch_txtbox,
- maxpitch_slider, maxpitch_txtbox]
- )
-
- with gr.Column():
- resample_sr0 = gr.Slider(
- minimum=0,
- maximum=48000,
- label=i18n("Resample the output audio in post-processing to the final sample rate. Set to 0 for no resampling:"),
- value=0,
- step=1,
- interactive=True,
- )
- rms_mix_rate0 = gr.Slider(
- minimum=0,
- maximum=1,
- label=i18n("Use the volume envelope of the input to replace or mix with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is used:"),
- value=0.25,
- interactive=True,
- )
- protect0 = gr.Slider(
- minimum=0,
- maximum=0.5,
- label=i18n(
- "Protect voiceless consonants and breath sounds to prevent artifacts such as tearing in electronic music. Set to 0.5 to disable. Decrease the value to increase protection, but it may reduce indexing accuracy:"
- ),
- value=0.33,
- step=0.01,
- interactive=True,
- )
- formanting = gr.Checkbox(
- value=bool(DoFormant),
- label=i18n("Formant shift inference audio"),
- info=i18n("Used for male to female and vice-versa conversions"),
- interactive=True,
- visible=True,
- )
-
- formant_preset = gr.Dropdown(
- value='',
- choices=get_fshift_presets(),
- label=i18n("Browse presets for formanting"),
- info=i18n("Presets are located in formantshiftcfg/ folder"),
- visible=bool(DoFormant),
- )
-
- formant_refresh_button = gr.Button(
- value='\U0001f504',
- visible=bool(DoFormant),
- variant='primary',
- )
-
- qfrency = gr.Slider(
- value=Quefrency,
- info=i18n("Default value is 1.0"),
- label=i18n("Quefrency for formant shifting"),
- minimum=0.0,
- maximum=16.0,
- step=0.1,
- visible=bool(DoFormant),
- interactive=True,
- )
-
- tmbre = gr.Slider(
- value=Timbre,
- info=i18n("Default value is 1.0"),
- label=i18n("Timbre for formant shifting"),
- minimum=0.0,
- maximum=16.0,
- step=0.1,
- visible=bool(DoFormant),
- interactive=True,
- )
- frmntbut = gr.Button(
- "Apply", variant="primary", visible=bool(DoFormant)
- )
-
- formant_preset.change(
- fn=preset_apply,
- inputs=[formant_preset, qfrency, tmbre],
- outputs=[qfrency, tmbre],
- )
- formanting.change(
- fn=formant_enabled,
- inputs=[
- formanting,
- qfrency,
- tmbre,
- frmntbut,
- formant_preset,
- formant_refresh_button,
- ],
- outputs=[
- formanting,
- qfrency,
- tmbre,
- frmntbut,
- formant_preset,
- formant_refresh_button,
- ],
- )
- frmntbut.click(
- fn=formant_apply,
- inputs=[qfrency, tmbre],
- outputs=[qfrency, tmbre],
- )
- formant_refresh_button.click(
- fn=update_fshift_presets,
- inputs=[formant_preset, qfrency, tmbre],
- outputs=[formant_preset, qfrency, tmbre],
- )
-
- # Function to toggle advanced settings
- def toggle_advanced_settings(checkbox):
- return {"visible": checkbox, "__type__": "update"}
-
- # Attach the change event
- advanced_settings_checkbox.change(
- fn=toggle_advanced_settings,
- inputs=[advanced_settings_checkbox],
- outputs=[advanced_settings]
- )
-
-
- but0 = gr.Button(i18n("Convert"), variant="primary").style(full_width=True)
-
- with gr.Row(): # Defines output info + output audio download after conversion
- vc_output1 = gr.Textbox(label=i18n("Output information:"))
- vc_output2 = gr.Audio(label=i18n("Export audio (click on the three dots in the lower right corner to download)"))
-
- with gr.Group(): # I think this defines the big convert button
- with gr.Row():
- but0.click(
- vc.vc_single,
- [
- spk_item,
- input_audio0,
- input_audio1,
- vc_transform0,
- f0_file,
- f0method0,
- file_index1,
- file_index2,
- index_rate1,
- filter_radius0,
- resample_sr0,
- rms_mix_rate0,
- protect0,
- crepe_hop_length,
- minpitch_slider, minpitch_txtbox,
- maxpitch_slider, maxpitch_txtbox,
- f0_autotune
- ],
- [vc_output1, vc_output2],
- )
-
-
- with gr.TabItem(i18n("Batch")): # Dont Change
- with gr.Group(): # Markdown explanation of batch inference
- gr.Markdown(
- value=i18n("Batch conversion. Enter the folder containing the audio files to be converted or upload multiple audio files. The converted audio will be output in the specified folder (default: 'opt').")
- )
- with gr.Row():
- with gr.Column():
- vc_transform1 = gr.Number(
- label=i18n("Transpose (integer, number of semitones, raise by an octave: 12, lower by an octave: -12):"), value=0
- )
- opt_input = gr.Textbox(label=i18n("Specify output folder:"), value="opt")
- with gr.Column():
- file_index4 = gr.Dropdown(
- label=i18n("Auto-detect index path and select from the dropdown:"),
- choices=get_indexes(),
- value=best_match_index_path1,
- interactive=True,
- )
- sid0.select(fn=match_index, inputs=[sid0], outputs=[file_index2, file_index4])
-
- refresh_button.click(
- fn=lambda: change_choices()[1],
- inputs=[],
- outputs=file_index4,
- )
- index_rate2 = gr.Slider(
- minimum=0,
- maximum=1,
- label=i18n("Search feature ratio:"),
- value=0.75,
- interactive=True,
- )
- with gr.Row():
- dir_input = gr.Textbox(
- label=i18n("Enter the path of the audio folder to be processed (copy it from the address bar of the file manager):"),
- value=os.path.join(now_dir, "audios"),
- )
- inputs = gr.File(
- file_count="multiple", label=i18n("You can also input audio files in batches. Choose one of the two options. Priority is given to reading from the folder.")
- )
-
- with gr.Row():
- with gr.Column():
- # Create a checkbox for advanced batch settings
- advanced_settings_batch_checkbox = gr.Checkbox(
- value=False,
- label=i18n("Advanced Settings"),
- interactive=True,
- )
-
- # Advanced batch settings container
- with gr.Row(visible=False) as advanced_settings_batch: # Initially hidden
- with gr.Row(label = i18n("Advanced Settings"), open = False):
- with gr.Column():
- file_index3 = gr.Textbox(
- label=i18n("Feature search database file path:"),
- value="",
- interactive=True,
- )
-
- f0method1 = gr.Radio(
- label=i18n(
- "Select the pitch extraction algorithm:"
- ),
- choices=["pm", "harvest", "crepe", "rmvpe"],
- value="rmvpe",
- interactive=True,
- )
- f0_autotune = gr.Checkbox(
- label="Enable autotune",
- interactive=True
- )
- filter_radius1 = gr.Slider(
- minimum=0,
- maximum=7,
- label=i18n("If >=3: apply median filtering to the harvested pitch results. The value represents the filter radius and can reduce breathiness."),
- value=3,
- step=1,
- interactive=True,
- )
-
- with gr.Row():
- format1 = gr.Radio(
- label=i18n("Export file format"),
- choices=["wav", "flac", "mp3", "m4a"],
- value="wav",
- interactive=True,
- )
-
-
- with gr.Column():
- resample_sr1 = gr.Slider(
- minimum=0,
- maximum=48000,
- label=i18n("Resample the output audio in post-processing to the final sample rate. Set to 0 for no resampling:"),
- value=0,
- step=1,
- interactive=True,
- )
- rms_mix_rate1 = gr.Slider(
- minimum=0,
- maximum=1,
- label=i18n("Use the volume envelope of the input to replace or mix with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is used:"),
- value=1,
- interactive=True,
- )
- protect1 = gr.Slider(
- minimum=0,
- maximum=0.5,
- label=i18n(
- "Protect voiceless consonants and breath sounds to prevent artifacts such as tearing in electronic music. Set to 0.5 to disable. Decrease the value to increase protection, but it may reduce indexing accuracy:"
- ),
- value=0.33,
- step=0.01,
- interactive=True,
- )
- vc_output3 = gr.Textbox(label=i18n("Output information:"))
- but1 = gr.Button(i18n("Convert"), variant="primary")
- but1.click(
- vc.vc_multi,
- [
- spk_item,
- dir_input,
- opt_input,
- inputs,
- vc_transform1,
- f0method1,
- file_index3,
- file_index4,
- index_rate2,
- filter_radius1,
- resample_sr1,
- rms_mix_rate1,
- protect1,
- format1,
- crepe_hop_length,
- minpitch_slider if (not rvc_globals.NotesOrHertz) else minpitch_txtbox,
- maxpitch_slider if (not rvc_globals.NotesOrHertz) else maxpitch_txtbox,
- f0_autotune
- ],
- [vc_output3],
- )
-
- sid0.change(
- fn=vc.get_vc,
- inputs=[sid0, protect0, protect1],
- outputs=[spk_item, protect0, protect1],
- )
- if not sid0.value == '':
- spk_item, protect0, protect1 = vc.get_vc(sid0.value, protect0, protect1)
-
- #spk_item, protect0, protect1 = vc.get_vc(sid0.value, protect0, protect1)
-
- # Function to toggle advanced settings
- def toggle_advanced_settings_batch(checkbox):
- return {"visible": checkbox, "__type__": "update"}
-
- # Attach the change event
- advanced_settings_batch_checkbox.change(
- fn=toggle_advanced_settings_batch,
- inputs=[advanced_settings_batch_checkbox],
- outputs=[advanced_settings_batch]
- )
-
-
- with gr.TabItem(i18n("Train")):
-
-
- with gr.Accordion(label=i18n("Step 1: Processing data")):
- with gr.Row():
- exp_dir1 = gr.Textbox(label=i18n("Enter the model name:"), value=i18n("Model_Name"))
- sr2 = gr.Radio(
- label=i18n("Target sample rate:"),
- choices=["40k", "48k", "32k"],
- value="40k",
- interactive=True,
- )
- if_f0_3 = gr.Checkbox(
- label=i18n("Whether the model has pitch guidance."),
- value=True,
- interactive=True,
- )
- version19 = gr.Radio(
- label=i18n("Version:"),
- choices=["v1", "v2"],
- value="v2",
- interactive=True,
- visible=True,
- )
- np7 = gr.Slider(
- minimum=0,
- maximum=config.n_cpu,
- step=1,
- label=i18n("Number of CPU processes:"),
- value=int(np.ceil(config.n_cpu / 1.5)),
- interactive=True,
- )
- with gr.Group():
- with gr.Accordion(label=i18n("Step 2: Skipping pitch extraction")):
-
- with gr.Row():
- # trainset_dir4 = gr.Textbox(
- # label=i18n("Enter the path of the training folder:"), value=os.path.join(now_dir, datasets_root)
- # )
- with gr.Column():
- trainset_dir4 = gr.Dropdown(choices=sorted(datasets), label=i18n("Select your dataset:"), value=get_dataset())
- btn_update_dataset_list = gr.Button(i18n("Update list"), variant="primary")
- spk_id5 = gr.Slider(
- minimum=0,
- maximum=4,
- step=1,
- label=i18n("Specify the model ID:"),
- value=0,
- interactive=True,
- )
- btn_update_dataset_list.click(
- easy_infer.update_dataset_list, [spk_id5], trainset_dir4
- )
- but1 = gr.Button(i18n("Process data"), variant="primary")
- info1 = gr.Textbox(label=i18n("Output information:"), value="")
- but1.click(
- preprocess_dataset, [trainset_dir4, exp_dir1, sr2, np7], [info1]
- )
- with gr.Group():
- with gr.Accordion(label=i18n("Step 3: Extracting features")):
- with gr.Row():
- with gr.Column():
- gpus6 = gr.Textbox(
- label=i18n("Provide the GPU index(es) separated by '-', like 0-1-2 for using GPUs 0, 1, and 2:"),
- value=gpus,
- interactive=True,
- )
- gpu_info9 = gr.Textbox(
- label=i18n("GPU Information:"), value=gpu_info, visible=F0GPUVisible
- )
- with gr.Column():
- f0method8 = gr.Radio(
- label=i18n(
- "Select the pitch extraction algorithm:"
- ),
- choices=["pm", "harvest", "dio", "crepe", "mangio-crepe", "rmvpe", "rmvpe_gpu"],
- # [ MANGIO ]: Fork feature: Crepe on f0 extraction for training.
- value="rmvpe",
- interactive=True,
- )
- gpus_rmvpe = gr.Textbox(
- label=i18n(
- "rmvpe卡号配置：以-分隔输入使用的不同进程卡号,例如0-0-1使用在卡0上跑2个进程并在卡1上跑1个进程"
- ),
- value="%s-%s" % (gpus, gpus),
- interactive=True,
- visible=F0GPUVisible,
- )
-
- extraction_crepe_hop_length = gr.Slider(
- minimum=1,
- maximum=512,
- step=1,
- label=i18n("Mangio-Crepe Hop Length (Only applies to mangio-crepe): Hop length refers to the time it takes for the speaker to jump to a dramatic pitch. Lower hop lengths take more time to infer but are more pitch accurate."),
- value=64,
- interactive=True,
- visible=False,
- )
-
- f0method8.change(
- fn=lambda radio: (
- {
- "visible": radio in ['mangio-crepe', 'mangio-crepe-tiny'],
- "__type__": "update"
- }
- ),
- inputs=[f0method8],
- outputs=[extraction_crepe_hop_length]
- )
- f0method8.change(
- fn=change_f0_method,
- inputs=[f0method8],
- outputs=[gpus_rmvpe],
- )
- but2 = gr.Button(i18n("Feature extraction"), variant="primary")
- info2 = gr.Textbox(label=i18n("Output information:"), value="", max_lines=8, interactive=False)
- but2.click(
- extract_f0_feature,
- [gpus6, np7, f0method8, if_f0_3, exp_dir1, version19, extraction_crepe_hop_length, gpus_rmvpe,],
- [info2],
- )
- with gr.Group():
- with gr.Row():
- with gr.Accordion(label=i18n("Step 4: Model training started")):
- with gr.Row():
- save_epoch10 = gr.Slider(
- minimum=1,
- maximum=100,
- step=1,
- label=i18n("Save frequency:"),
- value=10,
- interactive=True,
- visible=True,
- )
- total_epoch11 = gr.Slider(
- minimum=1,
- maximum=10000,
- step=2,
- label=i18n("Training epochs:"),
- value=750,
- interactive=True,
- )
- batch_size12 = gr.Slider(
- minimum=1,
- maximum=50,
- step=1,
- label=i18n("Batch size per GPU:"),
- value=default_batch_size,
- #value=20,
- interactive=True,
- )
-
- with gr.Row():
- if_save_latest13 = gr.Checkbox(
- label=i18n("Whether to save only the latest .ckpt file to save hard drive space"),
- value=True,
- interactive=True,
- )
- if_cache_gpu17 = gr.Checkbox(
- label=i18n("Cache all training sets to GPU memory. Caching small datasets (less than 10 minutes) can speed up training"),
- value=False,
- interactive=True,
- )
- if_save_every_weights18 = gr.Checkbox(
- label=i18n("Save a small final model to the 'weights' folder at each save point"),
- value=True,
- interactive=True,
- )
-
- with gr.Row():
- pretrained_G14 = gr.Textbox(
- lines=4,
- label=i18n("Load pre-trained base model G path:"),
- value="assets/pretrained_v2/f0G40k.pth",
- interactive=True,
- )
- pretrained_D15 = gr.Textbox(
- lines=4,
- label=i18n("Load pre-trained base model D path:"),
- value="assets/pretrained_v2/f0D40k.pth",
- interactive=True,
- )
- gpus16 = gr.Textbox(
- label=i18n("Provide the GPU index(es) separated by '-', like 0-1-2 for using GPUs 0, 1, and 2:"),
- value=gpus,
- interactive=True,
- )
- sr2.change(
- change_sr2,
- [sr2, if_f0_3, version19],
- [pretrained_G14, pretrained_D15],
- )
- version19.change(
- change_version19,
- [sr2, if_f0_3, version19],
- [pretrained_G14, pretrained_D15, sr2],
- )
- if_f0_3.change(
- fn=change_f0,
- inputs=[if_f0_3, sr2, version19],
- outputs=[f0method8, pretrained_G14, pretrained_D15],
- )
- if_f0_3.change(fn=lambda radio: (
- {
- "visible": radio in ['mangio-crepe', 'mangio-crepe-tiny'],
- "__type__": "update"
- }
- ), inputs=[f0method8], outputs=[extraction_crepe_hop_length])
-
- butstop = gr.Button(i18n("Stop training"),
- variant='primary',
- visible=False,
- )
- but3 = gr.Button(i18n("Train model"), variant="primary", visible=True)
- but3.click(fn=stoptraining, inputs=[gr.Number(value=0, visible=False)], outputs=[but3, butstop])
- butstop.click(fn=stoptraining, inputs=[gr.Number(value=1, visible=False)], outputs=[but3, butstop])
-
-
- with gr.Column():
- info3 = gr.Textbox(label=i18n("Output information:"), value="", max_lines=4)
- save_action = gr.Dropdown(label=i18n("Save type"), choices=[i18n("Save all"),i18n("Save D and G"),i18n("Save voice")], value=i18n("Choose the method"), interactive=True)
-
- but7 = gr.Button(i18n("Save model"), variant="primary")
- but4 = gr.Button(i18n("Train feature index"), variant="primary")
-
-
-
- if_save_every_weights18.change(
- fn=lambda if_save_every_weights: (
- {
- "visible": if_save_every_weights,
- "__type__": "update"
- }
- ),
- inputs=[if_save_every_weights18],
- outputs=[save_epoch10]
- )
-
- but3.click(
- click_train,
- [
- exp_dir1,
- sr2,
- if_f0_3,
- spk_id5,
- save_epoch10,
- total_epoch11,
- batch_size12,
- if_save_latest13,
- pretrained_G14,
- pretrained_D15,
- gpus16,
- if_cache_gpu17,
- if_save_every_weights18,
- version19,
- ],
- [info3, butstop, but3],
- )
-
- but4.click(train_index, [exp_dir1, version19], info3)
- but7.click(easy_infer.save_model, [exp_dir1, save_action], info3)
- with gr.Group():
- with gr.Row():
- with gr.Accordion(label=i18n("Step 5: Export lowest points on a graph of the model")):
-
- lowestval_weight_dir = gr.Textbox(visible=False)
- ds = gr.Textbox(visible=False)
- weights_dir1 = gr.Textbox(visible=False, value=weights_dir)
-
-
- with gr.Row():
- amntlastmdls = gr.Slider(
- minimum=1,
- maximum=25,
- label=i18n('How many lowest points to save:'),
- value=3,
- step=1,
- interactive=True,
- )
- lpexport = gr.Button(
- value=i18n('Export lowest points of a model'),
- variant='primary',
- )
- lw_mdls = gr.File(
- file_count="multiple",
- label=i18n("Output models:"),
- interactive=False,
- ) #####
-
- with gr.Row():
- infolpex = gr.Textbox(label=i18n("Output information:"), value="", max_lines=10)
- mdlbl = gr.Dataframe(label=i18n('Stats of selected models:'), datatype='number', type='pandas')
-
- lpexport.click(
- lambda model_name: os.path.join("logs", model_name, "lowestvals"),
- inputs=[exp_dir1],
- outputs=[lowestval_weight_dir]
- )
-
- lpexport.click(fn=tensorlowest.main, inputs=[exp_dir1, save_epoch10, amntlastmdls], outputs=[ds])
-
- ds.change(
- fn=tensorlowest.selectweights,
- inputs=[exp_dir1, ds, weights_dir1, lowestval_weight_dir],
- outputs=[infolpex, lw_mdls, mdlbl],
- )
- with gr.TabItem(i18n("UVR5")): # UVR section
- with gr.Group():
- with gr.Row():
- with gr.Column():
- model_select = gr.Radio(
- label=i18n("Model Architecture:"),
- choices=["VR", "MDX"],
- value="VR",
- interactive=True,
- )
- dir_wav_input = gr.Textbox(
- label=i18n("Enter the path of the audio folder to be processed:"),
- value=os.path.join(now_dir, "audios")
- )
- wav_inputs = gr.File(
- file_count="multiple", label=i18n("You can also input audio files in batches. Choose one of the two options. Priority is given to reading from the folder.")
- )
-
- with gr.Column():
- model_choose = gr.Dropdown(label=i18n("Model:"), choices=uvr5_names)
- agg = gr.Slider(
- minimum=0,
- maximum=20,
- step=1,
- label="Vocal Extraction Aggressive",
- value=10,
- interactive=True,
- visible=False,
- )
- opt_vocal_root = gr.Textbox(
- label=i18n("Specify the output folder for vocals:"), value="opt"
- )
- opt_ins_root = gr.Textbox(
- label=i18n("Specify the output folder for accompaniment:"), value="opt"
- )
- format0 = gr.Radio(
- label=i18n("Export file format:"),
- choices=["wav", "flac", "mp3", "m4a"],
- value="flac",
- interactive=True,
- )
- model_select.change(
- fn=update_model_choices,
- inputs=model_select,
- outputs=model_choose,
- )
- but2 = gr.Button(i18n("Convert"), variant="primary")
- vc_output4 = gr.Textbox(label=i18n("Output information:"))
- #wav_inputs.upload(fn=save_to_wav2_edited, inputs=[wav_inputs], outputs=[])
- but2.click(
- uvr,
- [
- model_choose,
- dir_wav_input,
- opt_vocal_root,
- wav_inputs,
- opt_ins_root,
- agg,
- format0,
- model_select
- ],
- [vc_output4],
- )
- with gr.TabItem(i18n("TTS")):
- with gr.Group():
- with gr.Column():
- text_test = gr.Textbox(label=i18n("Text:"), placeholder=i18n("Enter the text you want to convert to voice..."), lines=6)
-
- with gr.Group():
- with gr.Row():
- with gr.Column():
- tts_methods_voice = ["Edge-tts", "Bark-tts"]
- ttsmethod_test = gr.Dropdown(tts_methods_voice, value='Edge-tts', label = i18n('TTS Method:'), visible=True)
- tts_test = gr.Dropdown(set_edge_voice, label = i18n('TTS Model:'), visible=True)
- ttsmethod_test.change(
- fn=update_tts_methods_voice,
- inputs=ttsmethod_test,
- outputs=tts_test,
- )
-
- with gr.Column():
- model_voice_path07 = gr.Dropdown(label=i18n('RVC Model:'), choices=sorted(names), value=default_weight)
- best_match_index_path1 = match_index(model_voice_path07.value)
-
- file_index2_07 = gr.Dropdown(
- label=i18n('Select the .index file:'),
- choices=get_indexes(),
- value=best_match_index_path1,
- interactive=True,
- allow_custom_value=True,
- )
- #transpose_test = gr.Number(label = i18n('Transpose (integer, number Fof semitones, raise by an octave: 12, lower by an octave: -12):'), value=0, visible=True, interactive= True)
-
-
-
-
- with gr.Row():
- refresh_button_ = gr.Button(i18n("Refresh"), variant="primary")
- refresh_button_.click(fn=change_choices2, inputs=[], outputs=[model_voice_path07, file_index2_07])
- with gr.Row():
- original_ttsvoice = gr.Audio(label=i18n('Audio TTS:'))
- ttsvoice = gr.Audio(label=i18n('Audio RVC:'))
-
- with gr.Row():
- button_test = gr.Button(i18n("Convert"), variant="primary")
-
-
- button_test.click(make_test, inputs=[
- text_test,
- tts_test,
- model_voice_path07,
- file_index2_07,
- #transpose_test,
- vc_transform0,
- f0method8,
- index_rate1,
- crepe_hop_length,
- f0_autotune,
- ttsmethod_test
- ], outputs=[ttsvoice, original_ttsvoice])
-
- with gr.TabItem(i18n("Resources")):
- gr.Markdown(f"Limit download size is {os.getenv('MAX_DOWNLOAD_SIZE')} MB, duplicate the space for modify the limit")
- easy_infer.download_model()
- easy_infer.download_backup()
- easy_infer.download_dataset(trainset_dir4)
- easy_infer.download_audio()
- easy_infer.youtube_separator()
- with gr.TabItem(i18n("Extra")):
- gr.Markdown(
- value=i18n("This section contains some extra utilities that often may be in experimental phases")
- )
- with gr.TabItem(i18n("Merge Audios")):
- with gr.Group():
- gr.Markdown(
- value="## " + i18n("Merge your generated audios with the instrumental")
- )
- gr.Markdown(value=".",visible=True)
- gr.Markdown(value=".",visible=True)
- with gr.Row():
- with gr.Column():
- dropbox = gr.File(label=i18n("Drag your audio here:"))
- gr.Markdown(value=i18n("### Instrumental settings:"))
- input_audio1 = gr.Dropdown(
- label=i18n("Choose your instrumental:"),
- choices=sorted(audio_others_paths),
- value='',
- interactive=True,
- )
- input_audio1_scale = gr.Slider(
- minimum=0,
- maximum=10,
- label=i18n("Volume of the instrumental audio:"),
- value=1.00,
- interactive=True,
- )
- gr.Markdown(value=i18n("### Audio settings:"))
- input_audio3 = gr.Dropdown(
- label=i18n("Select the generated audio"),
- choices=sorted(audio_paths),
- value='',
- interactive=True,
- )
- with gr.Row():
- input_audio3_scale = gr.Slider(
- minimum=0,
- maximum=10,
- label=i18n("Volume of the generated audio:"),
- value=1.00,
- interactive=True,
- )
-
- gr.Markdown(value=i18n("### Add the effects:"))
- reverb_ = gr.Checkbox(
- label=i18n("Reverb"),
- value=False,
- interactive=True,
- )
- compressor_ = gr.Checkbox(
- label=i18n("Compressor"),
- value=False,
- interactive=True,
- )
- noise_gate_ = gr.Checkbox(
- label=i18n("Noise Gate"),
- value=False,
- interactive=True,
- )
-
- butnone = gr.Button(i18n("Merge"), variant="primary").style(full_width=True)
-
- vc_output1 = gr.Textbox(label=i18n("Output information:"))
- vc_output2 = gr.Audio(label=i18n("Export audio (click on the three dots in the lower right corner to download)"), type='filepath')
-
- dropbox.upload(fn=save_to_wav2, inputs=[dropbox], outputs=[input_audio1])
- dropbox.upload(fn=easy_infer.change_choices2, inputs=[], outputs=[input_audio1])
-
- refresh_button.click(
- fn=lambda: change_choices3(),
- inputs=[],
- outputs=[input_audio1, input_audio3],
- )
-
- butnone.click(
- fn=audio_combined,
- inputs=[input_audio1, input_audio3,input_audio1_scale,input_audio3_scale,reverb_,compressor_,noise_gate_],
- outputs=[vc_output1, vc_output2]
- )
-
-
- with gr.TabItem(i18n("Processing")):
- with gr.Group():
-
- with gr.Accordion(label=i18n("Model fusion, can be used to test timbre fusion")):
- with gr.Row():
- with gr.Column():
- name_to_save0 = gr.Textbox(
- label=i18n("Name:"),
- value="",
- max_lines=1,
- interactive=True,
- placeholder=i18n("Name for saving")
- )
- alpha_a = gr.Slider(
- minimum=0,
- maximum=1,
- label=i18n("Weight for Model A:"),
- value=0.5,
- interactive=True,
- )
- if_f0_ = gr.Checkbox(
- label=i18n("Whether the model has pitch guidance."),
- value=True,
- interactive=True,
- )
- version_2 = gr.Radio(
- label=i18n("Model architecture version:"),
- choices=["v1", "v2"],
- value="v2",
- interactive=True,
- )
- sr_ = gr.Radio(
- label=i18n("Target sample rate:"),
- choices=["40k", "48k"],
- value="40k",
- interactive=True,
- )
-
-
- with gr.Column():
- ckpt_a = gr.Textbox(label=i18n("Path to Model A:"), value="", interactive=True, placeholder=i18n("Path to model"))
-
- ckpt_b = gr.Textbox(label=i18n("Path to Model B:"), value="", interactive=True, placeholder=i18n("Path to model"))
-
- info__ = gr.Textbox(
- label=i18n("Model information to be placed:"), value="", max_lines=8, interactive=True, placeholder=i18n("Model information to be placed")
- )
- info4 = gr.Textbox(label=i18n("Output information:"), value="", max_lines=8)
-
-
- but6 = gr.Button(i18n("Fusion"), variant="primary")
-
- but6.click(
- merge,
- [
- ckpt_a,
- ckpt_b,
- alpha_a,
- sr_,
- if_f0_,
- info__,
- name_to_save0,
- version_2,
- ],
- info4,
- ) # def merge(path1,path2,alpha1,sr,f0,info):
- with gr.Group():
- with gr.Accordion(label=i18n("Modify model information")):
- with gr.Row(): ######
- with gr.Column():
- ckpt_path0 = gr.Textbox(
- label=i18n("Path to Model:"), value="", interactive=True, placeholder=i18n("Path to model")
- )
- info_ = gr.Textbox(
- label=i18n("Model information to be modified:"), value="", max_lines=8, interactive=True, placeholder=i18n("Model information to be placed")
- )
-
- with gr.Column():
- name_to_save1 = gr.Textbox(
- label=i18n("Save file name:"),
- placeholder=i18n("Name for saving"),
- value="",
- max_lines=8,
- interactive=True,
-
- )
-
- info5 = gr.Textbox(label=i18n("Output information:"), value="", max_lines=8)
- but7 = gr.Button(i18n("Modify"), variant="primary")
- but7.click(change_info, [ckpt_path0, info_, name_to_save1], info5)
- with gr.Group():
- with gr.Accordion(label=i18n("View model information")):
- with gr.Row():
- with gr.Column():
- ckpt_path1 = gr.Textbox(
- label=i18n("Path to Model:"), value="", interactive=True, placeholder=i18n("Path to model")
- )
-
- info6 = gr.Textbox(label=i18n("Output information:"), value="", max_lines=8)
- but8 = gr.Button(i18n("View"), variant="primary")
- but8.click(show_info, [ckpt_path1], info6)
- with gr.Group():
- with gr.Accordion(label=i18n("Model extraction")):
- with gr.Row():
- with gr.Column():
- save_name = gr.Textbox(
- label=i18n("Name:"), value="", interactive=True, placeholder=i18n("Name for saving")
- )
- if_f0__ = gr.Checkbox(
- label=i18n("Whether the model has pitch guidance."),
- value=True,
- interactive=True,
- )
- version_1 = gr.Radio(
- label=i18n("Model architecture version:"),
- choices=["v1", "v2"],
- value="v2",
- interactive=True,
- )
- sr__ = gr.Radio(
- label=i18n("Target sample rate:"),
- choices=["32k", "40k", "48k"],
- value="40k",
- interactive=True,
- )
-
- with gr.Column():
- ckpt_path2 = gr.Textbox(
-
- label=i18n("Path to Model:"),
- placeholder=i18n("Path to model"),
- interactive=True,
- )
- info___ = gr.Textbox(
- label=i18n("Model information to be placed:"), value="", max_lines=8, interactive=True, placeholder=i18n("Model information to be placed")
- )
- info7 = gr.Textbox(label=i18n("Output information:"), value="", max_lines=8)
-
- with gr.Row():
-
- but9 = gr.Button(i18n("Extract"), variant="primary")
- ckpt_path2.change(
- change_info_, [ckpt_path2], [sr__, if_f0__, version_1]
- )
- but9.click(
- extract_small_model,
- [ckpt_path2, save_name, sr__, if_f0__, info___, version_1],
- info7,
- )
-
-
-
-
- with gr.TabItem(i18n("Settings")):
- with gr.Row():
- gr.Markdown(value=
- i18n("Pitch settings")
- )
- noteshertz = gr.Checkbox(
- label = i18n("Whether to use note names instead of their hertz value. E.G. [C5, D6] instead of [523.25, 1174.66]Hz"),
- value = rvc_globals.NotesOrHertz,
- interactive = True,
- )
-
- noteshertz.change(fn=lambda nhertz: rvc_globals.__setattr__('NotesOrHertz', nhertz), inputs=[noteshertz], outputs=[])
-
- noteshertz.change(
- fn=switch_pitch_controls,
- inputs=[f0method0],
- outputs=[
- minpitch_slider, minpitch_txtbox,
- maxpitch_slider, maxpitch_txtbox,]
- )
- return app
-
-def GradioRun(app):
- share_gradio_link = config.iscolab or config.paperspace
- concurrency_count = 511
- max_size = 1022
-
- if (
- config.iscolab or config.paperspace
- ):
- app.queue(concurrency_count=concurrency_count, max_size=max_size).launch(
- favicon_path="./images/icon.png",
- )
- else:
- app.queue(concurrency_count=concurrency_count, max_size=max_size).launch(
- favicon_path=".\images\icon.png",
- )
if __name__ == "__main__":
- if os.name == 'nt':
- print(i18n("Any ConnectionResetErrors post-conversion are irrelevant and purely visual; they can be ignored.\n"))
- app = GradioSetup(UTheme=config.grtheme)
- GradioRun(app)
\ No newline at end of file
+ launch_gradio()
\ No newline at end of file
diff --git a/assets/Applio.ipynb b/assets/Applio.ipynb
new file mode 100644
index 0000000000000000000000000000000000000000..5823f539c120d55598120676f86bca7bf777841f
--- /dev/null
+++ b/assets/Applio.ipynb
@@ -0,0 +1,450 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "ymhGfgFSR17k"
+ },
+ "source": [
+ "## **Applio**\n",
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.\n",
+ "\n",
+ "[Support](https://discord.gg/IAHispano) — [Discord Bot](https://discord.com/oauth2/authorize?client_id=1144714449563955302&permissions=1376674695271&scope=bot%20applications.commands) — [Find Voices](https://applio.org/models) — [GitHub](https://github.com/IAHispano/Applio)\n",
+ "\n",
+ " \n",
+ "\n",
+ "### **Credits**\n",
+ "- Encryption method: [Hina](https://github.com/hinabl)\n",
+ "- Extra section: [Poopmaster](https://github.com/poiqazwsx)\n",
+ "- Main development: [Applio Team](https://github.com/IAHispano)\n",
+ "\n",
+ " \n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "vtON700qokuQ"
+ },
+ "outputs": [],
+ "source": [
+ "# @title **Install Applio**\n",
+ "\n",
+ "import codecs\n",
+ "import time\n",
+ "import os\n",
+ "import csv\n",
+ "import shutil\n",
+ "import tarfile\n",
+ "import subprocess\n",
+ "from pathlib import Path\n",
+ "from datetime import datetime\n",
+ "\n",
+ "rot_47 = lambda encoded_text: \"\".join(\n",
+ " [\n",
+ " (\n",
+ " chr(\n",
+ " (ord(c) - (ord(\"a\") if c.islower() else ord(\"A\")) - 47) % 26\n",
+ " + (ord(\"a\") if c.islower() else ord(\"A\"))\n",
+ " )\n",
+ " if c.isalpha()\n",
+ " else c\n",
+ " )\n",
+ " for c in encoded_text\n",
+ " ]\n",
+ ")\n",
+ "\n",
+ "org_name = rot_47(\"Vkkgdj\")\n",
+ "new_name = rot_47(\"kmjbmvh_hg\")\n",
+ "uioawhd = rot_47(codecs.decode(\"pbbxa://oqbpcj.kwu/QIPqaxivw/Ixxtqw.oqb\", \"rot_13\"))\n",
+ "uyadwa = codecs.decode(\"ncc.cl\", \"rot_13\")\n",
+ "!git clone --depth 1 $uioawhd\n",
+ "!mv $org_name $new_name\n",
+ "%cd $new_name/\n",
+ "\n",
+ "from IPython.display import clear_output, Javascript\n",
+ "\n",
+ "clear_output()\n",
+ "\n",
+ "E = Exception\n",
+ "B = print\n",
+ "\n",
+ "\n",
+ "def vidal_setup(ForceIn):\n",
+ " L = \"Kikpm.ovm.bu\"\n",
+ " K = \"/content/\"\n",
+ " C = ForceIn\n",
+ "\n",
+ " def F():\n",
+ " print(\"Installing pip packages...\")\n",
+ " subprocess.check_call([\"pip\", \"install\", \"-r\", \"requirements.txt\", \"--quiet\"])\n",
+ "\n",
+ " A = K + rot_47(L)\n",
+ " G = K + rot_47(L)\n",
+ " D = \"/\"\n",
+ " if not os.path.exists(A):\n",
+ " M = os.path.dirname(A)\n",
+ " os.makedirs(M, exist_ok=True)\n",
+ " print(\"No cached install found..\")\n",
+ " try:\n",
+ " N = rot_47(\n",
+ " codecs.decode(\n",
+ " \"pbbxa://pcooqvonikm.kw/QIPqaxivw/Ixxtqw/zmawtdm/uiqv/Kwtij/Xvxcz.biz.oh\",\n",
+ " \"rot_13\",\n",
+ " )\n",
+ " )\n",
+ " subprocess.run([\"wget\", \"-O\", A, N])\n",
+ " print(\"Download completed successfully!\")\n",
+ " except E as H:\n",
+ " print(str(H))\n",
+ " if os.path.exists(A):\n",
+ " os.remove(A)\n",
+ " if Path(A).exists():\n",
+ " with tarfile.open(G, \"r:gz\") as I:\n",
+ " for J in I.getmembers():\n",
+ " O = os.path.join(D, J.name)\n",
+ " try:\n",
+ " I.extract(J, D)\n",
+ " except E as H:\n",
+ " print(\"Failed to extract a file\")\n",
+ " C = True\n",
+ " print(f\"Extraction of {G} to {D} completed.\")\n",
+ " if os.path.exists(A):\n",
+ " os.remove(A)\n",
+ " if C:\n",
+ " F()\n",
+ " C = False\n",
+ " else:\n",
+ " F()\n",
+ "\n",
+ "\n",
+ "vidal_setup(False)\n",
+ "clear_output()\n",
+ "print(\"Finished installing requirements!\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "-7cQtXouqpQi"
+ },
+ "outputs": [],
+ "source": [
+ "# @title **Start Applio**\n",
+ "# @markdown ### Just activate this in case the share link of the gradio dont work\n",
+ "import codecs\n",
+ "import threading\n",
+ "import urllib.request\n",
+ "import time\n",
+ "import ipywidgets as widgets\n",
+ "from IPython.display import display\n",
+ "!npm install -g localtunnel\n",
+ "new_name = rot_47(\"kmjbmvh_hg\")\n",
+ "%cd \"/content/program_ml\"\n",
+ "uyadwa = codecs.decode(\"ncc.cl\", \"rot_13\")\n",
+ "share_tunnel = False # @param {type:\"boolean\"}\n",
+ "def start_applio():\n",
+ " if share_tunnel:\n",
+ " !python $uyadwa --listen\n",
+ " else:\n",
+ " !python $uyadwa --listen --share\n",
+ "\n",
+ "%load_ext tensorboard\n",
+ "%reload_ext tensorboard\n",
+ "%tensorboard --logdir logs --bind_all\n",
+ "\n",
+ "if \"autobackups\" not in globals():\n",
+ " autobackups = False\n",
+ "\n",
+ "if autobackups:\n",
+ " thread = threading.Thread(target=backup_files)\n",
+ " thread.start()\n",
+ "\n",
+ "thread_applio = threading.Thread(target=start_applio)\n",
+ "thread_applio.start()\n",
+ "\n",
+ "if share_tunnel:\n",
+ " if not os.path.exists(codecs.decode(\"eip/cergenvarqf/cergenvarq_i2/s0T48x.cgu\", \"rot_13\")):\n",
+ " while not os.path.exists(codecs.decode(\"eip/cergenvarqf/cergenvarq_i2/s0T48x.cgu\", \"rot_13\")):\n",
+ " time.sleep(2)\n",
+ " time.sleep(5)\n",
+ " else:\n",
+ " time.sleep(10)\n",
+ " with open('url.txt', 'w') as file:\n",
+ " file.write('')\n",
+ "\n",
+ " get_ipython().system_raw('lt --port 6969 >> url.txt 2>&1 &')\n",
+ "\n",
+ " time.sleep(4)\n",
+ "\n",
+ " endpoint_ip = urllib.request.urlopen('https://ipv4.icanhazip.com').read().decode('utf8').strip(\"\\n\")\n",
+ "\n",
+ " with open('url.txt', 'r') as file:\n",
+ " tunnel_url = file.read()\n",
+ " tunnel_url = tunnel_url.replace(\"your url is: \", \"\")\n",
+ "\n",
+ " print(f\"Share Link: \\033[0m\\033[93m{tunnel_url}\\033[0m\", end=\"\\033[0m\\n\")\n",
+ "\n",
+ " password_endpoint_widget = widgets.Text(\n",
+ " value=endpoint_ip,\n",
+ " description='Password IP:',\n",
+ " disabled=True\n",
+ " )\n",
+ " display(password_endpoint_widget)\n",
+ "\n",
+ "\n",
+ "\n",
+ "while True:\n",
+ " time.sleep(5)\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "3b59-2x-qEnX"
+ },
+ "source": [
+ "### **Extra**\n",
+ "Enjoy extra options that can make it easier for you to use Applio\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "19LNv6iYqF6_"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Mount Drive\n",
+ "# @markdown Mount the files from Google Drive to the Colab.\n",
+ "from google.colab import drive\n",
+ "\n",
+ "drive.mount(\"/content/drive\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "I5o6MlpFouiG"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Auto Backup\n",
+ "# @markdown When running it, it will be activated or deactivated previously to start up together with Applio.\n",
+ "LOGS_FOLDER = \"/content/program_ml/logs/\"\n",
+ "GOOGLE_DRIVE_PATH = \"/content/drive/MyDrive/ApplioBackup\"\n",
+ "\n",
+ "if \"autobackups\" not in globals():\n",
+ " autobackups = False\n",
+ "\n",
+ "\n",
+ "def backup_files():\n",
+ " print(\"\\nStarting backup loop...\")\n",
+ " last_backup_timestamps_path = os.path.join(\n",
+ " LOGS_FOLDER, \"last_backup_timestamps.txt\"\n",
+ " )\n",
+ " fully_updated = False\n",
+ "\n",
+ " while True:\n",
+ " try:\n",
+ " updated = False\n",
+ " last_backup_timestamps = {}\n",
+ "\n",
+ " try:\n",
+ " with open(last_backup_timestamps_path, \"r\") as f:\n",
+ " last_backup_timestamps = dict(line.strip().split(\":\") for line in f)\n",
+ " except FileNotFoundError:\n",
+ " pass\n",
+ "\n",
+ " for root, dirs, files in os.walk(LOGS_FOLDER):\n",
+ " # Excluding \"zips\" directory\n",
+ " if \"zips\" in dirs:\n",
+ " dirs.remove(\"zips\")\n",
+ " if \"mute\" in dirs:\n",
+ " dirs.remove(\"mute\")\n",
+ " for filename in files:\n",
+ " if filename != \"last_backup_timestamps.txt\":\n",
+ " filepath = os.path.join(root, filename)\n",
+ " if os.path.isfile(filepath):\n",
+ " backup_filepath = os.path.join(\n",
+ " GOOGLE_DRIVE_PATH,\n",
+ " os.path.relpath(filepath, LOGS_FOLDER),\n",
+ " )\n",
+ " backup_folderpath = os.path.dirname(backup_filepath)\n",
+ " if not os.path.exists(backup_folderpath):\n",
+ " os.makedirs(backup_folderpath)\n",
+ " print(\n",
+ " f\"Created backup folder: {backup_folderpath}\",\n",
+ " flush=True,\n",
+ " )\n",
+ " last_backup_timestamp = last_backup_timestamps.get(filepath)\n",
+ " current_timestamp = os.path.getmtime(filepath)\n",
+ " if (\n",
+ " last_backup_timestamp is None\n",
+ " or float(last_backup_timestamp) < current_timestamp\n",
+ " ):\n",
+ " shutil.copy2(filepath, backup_filepath)\n",
+ " last_backup_timestamps[filepath] = str(\n",
+ " current_timestamp\n",
+ " )\n",
+ " if last_backup_timestamp is None:\n",
+ " print(f\"Backed up file: {filename}\")\n",
+ " else:\n",
+ " print(f\"Updating backed up file: {filename}\")\n",
+ " updated = True\n",
+ " fully_updated = False\n",
+ "\n",
+ " for filepath in list(last_backup_timestamps.keys()):\n",
+ " if not os.path.exists(filepath):\n",
+ " backup_filepath = os.path.join(\n",
+ " GOOGLE_DRIVE_PATH, os.path.relpath(filepath, LOGS_FOLDER)\n",
+ " )\n",
+ " if os.path.exists(backup_filepath):\n",
+ " os.remove(backup_filepath)\n",
+ " print(f\"Deleted file: {filepath}\")\n",
+ " del last_backup_timestamps[filepath]\n",
+ " updated = True\n",
+ " fully_updated = False\n",
+ "\n",
+ " if not updated and not fully_updated:\n",
+ " print(\"Files are up to date.\")\n",
+ " fully_updated = True\n",
+ " sleep_time = 15\n",
+ " else:\n",
+ " sleep_time = 0.1\n",
+ "\n",
+ " with open(last_backup_timestamps_path, \"w\") as f:\n",
+ " for filepath, timestamp in last_backup_timestamps.items():\n",
+ " f.write(f\"{filepath}:{timestamp}\\n\")\n",
+ "\n",
+ " time.sleep(sleep_time)\n",
+ "\n",
+ " except Exception as e:\n",
+ " print(f\"An error occurred: {str(e)}\")\n",
+ "\n",
+ "\n",
+ "if autobackups:\n",
+ " autobackups = False\n",
+ " print(\"Autobackup Disabled\")\n",
+ "else:\n",
+ " autobackups = True\n",
+ " print(\"Autobackup Enabled\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "ifV_vc4h4Uvx"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Load a Backup\n",
+ "from google.colab import drive\n",
+ "\n",
+ "# @markdown Put the exact name you put as your Model Name in Applio.\n",
+ "modelname = \"My-Project\" # @param {type:\"string\"}\n",
+ "source_path = \"/content/drive/MyDrive/ApplioBackup/\" + modelname\n",
+ "destination_path = \"/content/program_ml/logs/\" + modelname\n",
+ "backup_timestamps_file = \"last_backup_timestamps.txt\"\n",
+ "if not os.path.exists(source_path):\n",
+ " print(\n",
+ " \"The model folder does not exist. Please verify the name is correct or check your Google Drive.\"\n",
+ " )\n",
+ "else:\n",
+ " time_ = os.path.join(\"/content/drive/MyDrive/ApplioBackup/\", backup_timestamps_file)\n",
+ " time__ = os.path.join(\"/content/program_ml/logs/\", backup_timestamps_file)\n",
+ " if os.path.exists(time_):\n",
+ " shutil.copy(time_, time__)\n",
+ " shutil.copytree(source_path, destination_path)\n",
+ " print(\"Model backup loaded successfully.\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "leWbhk1X4XoY"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Download all custom pretrains\n",
+ "import os\n",
+ "import urllib.request\n",
+ "\n",
+ "%mkdir /content/program_ml/rvc/pretraineds/pretraineds_custom\n",
+ "pretrained_urls = [\n",
+ " # Ov2 Super\n",
+ " \"https://huggingface.co/ORVC/Ov2Super/resolve/main/f0Ov2Super32kG.pth\",\n",
+ " \"https://huggingface.co/ORVC/Ov2Super/resolve/main/f0Ov2Super32kD.pth\",\n",
+ " \"https://huggingface.co/ORVC/Ov2Super/resolve/main/f0Ov2Super40kG.pth\",\n",
+ " \"https://huggingface.co/ORVC/Ov2Super/resolve/main/f0Ov2Super40kD.pth\",\n",
+ "\n",
+ " # TITAN\n",
+ " \"https://huggingface.co/blaise-tk/TITAN/resolve/main/models/medium/40k/pretrained/G-f040k-TITAN-Medium.pth\",\n",
+ " \"https://huggingface.co/blaise-tk/TITAN/resolve/main/models/medium/40k/pretrained/D-f040k-TITAN-Medium.pth\",\n",
+ " \"https://huggingface.co/blaise-tk/TITAN/resolve/main/models/medium/32k/pretrained/G-f032k-TITAN-Medium.pth\",\n",
+ " \"https://huggingface.co/blaise-tk/TITAN/resolve/main/models/medium/32k/pretrained/D-f032k-TITAN-Medium.pth\",\n",
+ "\n",
+ " # Snowie V3\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-32k/resolve/main/D_SnowieV3.1_32k.pth\",\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-32k/resolve/main/G_SnowieV3.1_32k.pth\",\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-40k/resolve/main/G_SnowieV3.1_40k.pth\",\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-40k/resolve/main/D_SnowieV3.1_40k.pth\",\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-48k/resolve/main/G_SnowieV3.1_48k.pth\",\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-48k/resolve/main/D_SnowieV3.1_48k.pth\",\n",
+ "\n",
+ " # RIN E3\n",
+ " \"https://huggingface.co/MUSTAR/RIN_E3/resolve/main/RIN_E3_G.pth\",\n",
+ " \"https://huggingface.co/MUSTAR/RIN_E3/resolve/main/RIN_E3_D.pth\",\n",
+ "\n",
+ " # KLM\n",
+ " \"https://huggingface.co/SeoulStreamingStation/KLMv7s/resolve/main/G_KLMv7s_Batch2_32k.pth\",\n",
+ " \"https://huggingface.co/SeoulStreamingStation/KLMv7s/resolve/main/D_KLMv7s_Batch2_32k.pth\",\n",
+ " \"https://huggingface.co/SeoulStreamingStation/KLMv7s/resolve/main/G_KLMv7s_Batch2_40k.pth\",\n",
+ " \"https://huggingface.co/SeoulStreamingStation/KLMv7s/resolve/main/D_KLMv7s_Batch2_40k.pth\",\n",
+ " \"https://huggingface.co/SeoulStreamingStation/KLMv7s/resolve/main/G_KLMv7s_Batch2_48k.pth\",\n",
+ " \"https://huggingface.co/SeoulStreamingStation/KLMv7s/resolve/main/D_KLMv7s_Batch2_48k.pth\",\n",
+ "\n",
+ " # SnowieV3 X RIN_E3\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-X-RinE3-40K/resolve/main/D_Snowie-X-Rin_40k.pth\",\n",
+ " \"https://huggingface.co/MUSTAR/SnowieV3.1-X-RinE3-40K/resolve/main/G_Snowie-X-Rin_40k.pth\",\n",
+ "]\n",
+ "output_directory = \"/content/program_ml/rvc/pretraineds/pretraineds_custom\"\n",
+ "for url in pretrained_urls:\n",
+ " filename = os.path.join(output_directory, os.path.basename(url))\n",
+ " urllib.request.urlretrieve(url, filename)"
+ ]
+ }
+ ],
+ "metadata": {
+ "accelerator": "GPU",
+ "colab": {
+ "collapsed_sections": [
+ "3b59-2x-qEnX"
+ ],
+ "gpuType": "T4",
+ "provenance": []
+ },
+ "kernelspec": {
+ "display_name": "Python 3",
+ "name": "python3"
+ },
+ "language_info": {
+ "name": "python"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/assets/Applio_NoUI.ipynb b/assets/Applio_NoUI.ipynb
new file mode 100644
index 0000000000000000000000000000000000000000..c65d6b3a463643e68e2b94227863d7c4ef9f885a
--- /dev/null
+++ b/assets/Applio_NoUI.ipynb
@@ -0,0 +1,702 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "0pKllbPyK_BC"
+ },
+ "source": [
+ "# Applio NoUI\n",
+ "Created by [Blaise](https://github.com/blaise-tk) with [Vidal](https://github.com/Vidalnt) and [Poopmaster](https://github.com/poiqazwsx). Based on [RVC_CLI](https://github.com/blaise-tk/RVC_CLI).\n",
+ "\n",
+ "- Colab inspired on [RVC v2 Disconnected](https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS).\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "Y-iR3WeLMlac"
+ },
+ "source": [
+ "### If you restart the runtime, run it again."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "id": "xwZkZGd-H0zT"
+ },
+ "outputs": [],
+ "source": [
+ "%cd /content/Applio"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "ymMCTSD6m8qV"
+ },
+ "source": [
+ "# Installation\n",
+ "## If the runtime restarts, run the cell above and re-run the installation steps."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "yFhAeKGOp9aa"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Mount Google Drive\n",
+ "from google.colab import drive\n",
+ "\n",
+ "drive.mount(\"/content/drive\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "7GysECSxBya4"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Clone\n",
+ "!git clone https://github.com/IAHispano/Applio\n",
+ "%cd /content/Applio"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "CAXW55BQm0PP"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Install\n",
+ "rot_47 = lambda encoded_text: \"\".join(\n",
+ " [\n",
+ " (\n",
+ " chr(\n",
+ " (ord(c) - (ord(\"a\") if c.islower() else ord(\"A\")) - 47) % 26\n",
+ " + (ord(\"a\") if c.islower() else ord(\"A\"))\n",
+ " )\n",
+ " if c.isalpha()\n",
+ " else c\n",
+ " )\n",
+ " for c in encoded_text\n",
+ " ]\n",
+ ")\n",
+ "import codecs\n",
+ "import os\n",
+ "import shutil\n",
+ "import tarfile\n",
+ "import subprocess\n",
+ "from pathlib import Path\n",
+ "from datetime import datetime\n",
+ "E = Exception\n",
+ "B = print\n",
+ "\n",
+ "\n",
+ "def vidal_setup(ForceIn):\n",
+ " L = \"Kikpm.ovm.bu\"\n",
+ " K = \"/content/\"\n",
+ " C = ForceIn\n",
+ "\n",
+ " def F():\n",
+ " print(\"Installing pip packages...\")\n",
+ " subprocess.check_call([\"pip\", \"install\", \"-r\", \"requirements.txt\", \"--quiet\"])\n",
+ "\n",
+ " A = K + rot_47(L)\n",
+ " G = K + rot_47(L)\n",
+ " D = \"/\"\n",
+ " if not os.path.exists(A):\n",
+ " M = os.path.dirname(A)\n",
+ " os.makedirs(M, exist_ok=True)\n",
+ " print(\"No cached install found..\")\n",
+ " try:\n",
+ " N = rot_47(\n",
+ " codecs.decode(\n",
+ " \"pbbxa://pcooqvonikm.kw/QIPqaxivw/Ixxtqw/zmawtdm/uiqv/Kwtij/Xvxcz.biz.oh\",\n",
+ " \"rot_13\",\n",
+ " )\n",
+ " )\n",
+ " subprocess.run([\"wget\", \"-O\", A, N])\n",
+ " print(\"Download completed successfully!\")\n",
+ " except E as H:\n",
+ " print(str(H))\n",
+ " if os.path.exists(A):\n",
+ " os.remove(A)\n",
+ " if Path(A).exists():\n",
+ " with tarfile.open(G, \"r:gz\") as I:\n",
+ " for J in I.getmembers():\n",
+ " O = os.path.join(D, J.name)\n",
+ " try:\n",
+ " I.extract(J, D)\n",
+ " except E as H:\n",
+ " print(\"Failed to extract a file\")\n",
+ " C = True\n",
+ " print(f\"Extraction of {G} to {D} completed.\")\n",
+ " if os.path.exists(A):\n",
+ " os.remove(A)\n",
+ " if C:\n",
+ " F()\n",
+ " C = False\n",
+ " else:\n",
+ " F()\n",
+ "\n",
+ "\n",
+ "vidal_setup(False)\n",
+ "print(\"Finished installing requirements!\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "QlTibPnjmj6-"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Download models\n",
+ "!python core.py prerequisites"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "YzaeMYsUE97Y"
+ },
+ "source": [
+ "# Infer\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "v0EgikgjFCjE"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Download model\n",
+ "# @markdown Hugging Face or Google Drive\n",
+ "model_link = \"https://huggingface.co/Darwin/Darwin/resolve/main/Darwin.zip\" # @param {type:\"string\"}\n",
+ "\n",
+ "!python core.py download --model_link \"{model_link}\""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "lrCKEOzvDPRu"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Run Inference\n",
+ "# @markdown Please upload the audio file to your Google Drive path `/content/drive/MyDrive` and specify its name here. For the model name, use the zip file name without the extension. Alternatively, you can check the path `/content/Applio/logs` for the model name (name of the folder).\n",
+ "\n",
+ "import os\n",
+ "\n",
+ "current_dir = os.getcwd()\n",
+ "\n",
+ "model_name = \"Darwin\" # @param {type:\"string\"}\n",
+ "model_folder = os.path.join(current_dir, f\"logs/{model_name}\")\n",
+ "\n",
+ "if not os.path.exists(model_folder):\n",
+ " raise FileNotFoundError(f\"Model directory not found: {model_folder}\")\n",
+ "\n",
+ "files_in_folder = os.listdir(model_folder)\n",
+ "pth_path = next((f for f in files_in_folder if f.endswith(\".pth\")), None)\n",
+ "index_file = next((f for f in files_in_folder if f.endswith(\".index\")), None)\n",
+ "\n",
+ "if pth_path is None or index_file is None:\n",
+ " raise FileNotFoundError(\"No model found.\")\n",
+ "\n",
+ "pth_file = os.path.join(model_folder, pth_path)\n",
+ "index_file = os.path.join(model_folder, index_file)\n",
+ "\n",
+ "input_path = \"/content/example.wav\" # @param {type:\"string\"}\n",
+ "output_path = \"/content/output.wav\"\n",
+ "export_format = \"WAV\" # @param ['WAV', 'MP3', 'FLAC', 'OGG', 'M4A'] {allow-input: false}\n",
+ "f0method = \"rmvpe\" # @param [\"pm\", \"dio\", \"crepe\", \"crepe-tiny\", \"harvest\", \"rmvpe\", \"fcpe\", \"hybrid[rmvpe+fcpe]\"] {allow-input: false}\n",
+ "f0up_key = 0 # @param {type:\"slider\", min:-24, max:24, step:0}\n",
+ "filter_radius = 3 # @param {type:\"slider\", min:0, max:10, step:0}\n",
+ "rms_mix_rate = 0.8 # @param {type:\"slider\", min:0.0, max:1.0, step:0.1}\n",
+ "protect = 0.5 # @param {type:\"slider\", min:0.0, max:0.5, step:0.1}\n",
+ "index_rate = 0.7 # @param {type:\"slider\", min:0.0, max:1.0, step:0.1}\n",
+ "hop_length = 128 # @param {type:\"slider\", min:1, max:512, step:0}\n",
+ "clean_strength = 0.7 # @param {type:\"slider\", min:0.0, max:1.0, step:0.1}\n",
+ "split_audio = False # @param{type:\"boolean\"}\n",
+ "clean_audio = False # @param{type:\"boolean\"}\n",
+ "autotune = False # @param{type:\"boolean\"}\n",
+ "\n",
+ "!python core.py infer --f0up_key \"{f0up_key}\" --filter_radius \"{filter_radius}\" --index_rate \"{index_rate}\" --hop_length \"{hop_length}\" --rms_mix_rate \"{rms_mix_rate}\" --protect \"{protect}\" --f0autotune \"{autotune}\" --f0method \"{f0method}\" --input_path \"{input_path}\" --output_path \"{output_path}\" --pth_path \"{pth_file}\" --index_path \"{index_file}\" --split_audio \"{split_audio}\" --clean_audio \"{clean_audio}\" --clean_strength \"{clean_strength}\" --export_format \"{export_format}\"\n",
+ "\n",
+ "from IPython.display import Audio, display, clear_output\n",
+ "\n",
+ "output_path = output_path.replace(\".wav\", f\".{export_format.lower()}\")\n",
+ "# clear_output()\n",
+ "display(Audio(output_path, autoplay=True))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "1QkabnLlF2KB"
+ },
+ "source": [
+ "# Train"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "oBzqm4JkGGa0"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Preprocess Dataset\n",
+ "model_name = \"Darwin\" # @param {type:\"string\"}\n",
+ "dataset_path = \"/content/drive/MyDrive/Darwin_Dataset\" # @param {type:\"string\"}\n",
+ "\n",
+ "sample_rate = \"40k\" # @param [\"32k\", \"40k\", \"48k\"] {allow-input: false}\n",
+ "sr = int(sample_rate.rstrip(\"k\")) * 1000\n",
+ "\n",
+ "!python core.py preprocess --model_name \"{model_name}\" --dataset_path \"{dataset_path}\" --sampling_rate \"{sr}\""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "zWMiMYfRJTJv"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Extract Features\n",
+ "rvc_version = \"v2\" # @param [\"v2\", \"v1\"] {allow-input: false}\n",
+ "f0method = \"rmvpe\" # @param [\"pm\", \"dio\", \"crepe\", \"crepe-tiny\", \"harvest\", \"rmvpe\"] {allow-input: false}\n",
+ "hop_length = 128 # @param {type:\"slider\", min:1, max:512, step:0}\n",
+ "\n",
+ "sr = int(sample_rate.rstrip(\"k\")) * 1000\n",
+ "\n",
+ "!python core.py extract --model_name \"{model_name}\" --rvc_version \"{rvc_version}\" --f0method \"{f0method}\" --hop_length \"{hop_length}\" --sampling_rate \"{sr}\""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "TI6LLdIzKAIa"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Train\n",
+ "# @markdown ### ➡️ Model Information\n",
+ "import requests\n",
+ "import threading\n",
+ "import time\n",
+ "import os\n",
+ "import shutil\n",
+ "import hashlib\n",
+ "import time\n",
+ "\n",
+ "LOGS_FOLDER = \"/content/Applio/logs/\"\n",
+ "WEIGHTS_FOLDER = LOGS_FOLDER + model_name\n",
+ "GOOGLE_DRIVE_PATH = \"/content/drive/MyDrive/RVC_Backup\"\n",
+ "\n",
+ "\n",
+ "def import_google_drive_backup():\n",
+ " print(\"Importing Google Drive backup...\")\n",
+ " weights_exist = False\n",
+ " for root, dirs, files in os.walk(GOOGLE_DRIVE_PATH):\n",
+ " for filename in files:\n",
+ " filepath = os.path.join(root, filename)\n",
+ " if os.path.isfile(filepath) and not filepath.startswith(\n",
+ " os.path.join(GOOGLE_DRIVE_PATH, \"weights\")\n",
+ " ):\n",
+ " backup_filepath = os.path.join(\n",
+ " LOGS_FOLDER, os.path.relpath(filepath, GOOGLE_DRIVE_PATH)\n",
+ " )\n",
+ " backup_folderpath = os.path.dirname(backup_filepath)\n",
+ " if not os.path.exists(backup_folderpath):\n",
+ " os.makedirs(backup_folderpath)\n",
+ " print(f\"Created backup folder: {backup_folderpath}\", flush=True)\n",
+ " shutil.copy2(filepath, backup_filepath) # copy file with metadata\n",
+ " print(f\"Imported file from Google Drive backup: {filename}\")\n",
+ " elif filepath.startswith(\n",
+ " os.path.join(GOOGLE_DRIVE_PATH, \"weights\")\n",
+ " ) and filename.endswith(\".pth\"):\n",
+ " weights_exist = True\n",
+ " weights_filepath = os.path.join(\n",
+ " WEIGHTS_FOLDER,\n",
+ " os.path.relpath(\n",
+ " filepath, os.path.join(GOOGLE_DRIVE_PATH, \"weights\")\n",
+ " ),\n",
+ " )\n",
+ " weights_folderpath = os.path.dirname(weights_filepath)\n",
+ " if not os.path.exists(weights_folderpath):\n",
+ " os.makedirs(weights_folderpath)\n",
+ " print(f\"Created weights folder: {weights_folderpath}\", flush=True)\n",
+ " shutil.copy2(filepath, weights_filepath) # copy file with metadata\n",
+ " print(f\"Imported file from weights: {filename}\")\n",
+ " if weights_exist:\n",
+ " print(\"Copied weights from Google Drive backup to local weights folder.\")\n",
+ " else:\n",
+ " print(\"No weights found in Google Drive backup.\")\n",
+ " print(\"Google Drive backup import completed.\")\n",
+ "\n",
+ "\n",
+ "def get_md5_hash(file_path):\n",
+ " hash_md5 = hashlib.md5()\n",
+ " with open(file_path, \"rb\") as f:\n",
+ " for chunk in iter(lambda: f.read(4096), b\"\"):\n",
+ " hash_md5.update(chunk)\n",
+ " return hash_md5.hexdigest()\n",
+ "\n",
+ "\n",
+ "def copy_weights_folder_to_drive():\n",
+ " destination_folder = os.path.join(GOOGLE_DRIVE_PATH, \"weights\")\n",
+ " try:\n",
+ " if not os.path.exists(destination_folder):\n",
+ " os.makedirs(destination_folder)\n",
+ "\n",
+ " num_copied = 0\n",
+ " for filename in os.listdir(WEIGHTS_FOLDER):\n",
+ " if filename.endswith(\".pth\"):\n",
+ " source_file = os.path.join(WEIGHTS_FOLDER, filename)\n",
+ " destination_file = os.path.join(destination_folder, filename)\n",
+ " if not os.path.exists(destination_file):\n",
+ " shutil.copy2(source_file, destination_file)\n",
+ " num_copied += 1\n",
+ " print(f\"Copied {filename} to Google Drive!\")\n",
+ "\n",
+ " if num_copied == 0:\n",
+ " print(\"No new finished models found for copying.\")\n",
+ " else:\n",
+ " print(f\"Finished copying {num_copied} files to Google Drive!\")\n",
+ "\n",
+ " except Exception as e:\n",
+ " print(f\"An error occurred while copying weights: {str(e)}\")\n",
+ "\n",
+ "\n",
+ "if \"autobackups\" not in globals():\n",
+ " autobackups = False\n",
+ "\n",
+ "\n",
+ "def backup_files():\n",
+ " print(\"\\nStarting backup loop...\")\n",
+ " last_backup_timestamps_path = os.path.join(\n",
+ " LOGS_FOLDER, \"last_backup_timestamps.txt\"\n",
+ " )\n",
+ " fully_updated = False\n",
+ "\n",
+ " while True:\n",
+ " try:\n",
+ " updated = False\n",
+ " last_backup_timestamps = {}\n",
+ "\n",
+ " try:\n",
+ " with open(last_backup_timestamps_path, \"r\") as f:\n",
+ " last_backup_timestamps = dict(line.strip().split(\":\") for line in f)\n",
+ " except FileNotFoundError:\n",
+ " pass\n",
+ "\n",
+ " for root, dirs, files in os.walk(LOGS_FOLDER):\n",
+ " # Excluding \"zips\" directory\n",
+ " if \"zips\" in dirs:\n",
+ " dirs.remove(\"zips\")\n",
+ " if \"mute\" in dirs:\n",
+ " dirs.remove(\"mute\")\n",
+ " for filename in files:\n",
+ " if filename != \"last_backup_timestamps.txt\":\n",
+ " filepath = os.path.join(root, filename)\n",
+ " if os.path.isfile(filepath):\n",
+ " backup_filepath = os.path.join(\n",
+ " GOOGLE_DRIVE_PATH,\n",
+ " os.path.relpath(filepath, LOGS_FOLDER),\n",
+ " )\n",
+ " backup_folderpath = os.path.dirname(backup_filepath)\n",
+ " if not os.path.exists(backup_folderpath):\n",
+ " os.makedirs(backup_folderpath)\n",
+ " print(\n",
+ " f\"Created backup folder: {backup_folderpath}\",\n",
+ " flush=True,\n",
+ " )\n",
+ " last_backup_timestamp = last_backup_timestamps.get(filepath)\n",
+ " current_timestamp = os.path.getmtime(filepath)\n",
+ " if (\n",
+ " last_backup_timestamp is None\n",
+ " or float(last_backup_timestamp) < current_timestamp\n",
+ " ):\n",
+ " shutil.copy2(filepath, backup_filepath)\n",
+ " last_backup_timestamps[filepath] = str(\n",
+ " current_timestamp\n",
+ " )\n",
+ " if last_backup_timestamp is None:\n",
+ " print(f\"Backed up file: {filename}\")\n",
+ " else:\n",
+ " print(f\"Updating backed up file: {filename}\")\n",
+ " updated = True\n",
+ " fully_updated = False\n",
+ "\n",
+ " for filepath in list(last_backup_timestamps.keys()):\n",
+ " if not os.path.exists(filepath):\n",
+ " backup_filepath = os.path.join(\n",
+ " GOOGLE_DRIVE_PATH, os.path.relpath(filepath, LOGS_FOLDER)\n",
+ " )\n",
+ " if os.path.exists(backup_filepath):\n",
+ " os.remove(backup_filepath)\n",
+ " print(f\"Deleted file: {filepath}\")\n",
+ " del last_backup_timestamps[filepath]\n",
+ " updated = True\n",
+ " fully_updated = False\n",
+ "\n",
+ " if not updated and not fully_updated:\n",
+ " print(\"Files are up to date.\")\n",
+ " fully_updated = True\n",
+ " sleep_time = 15\n",
+ " else:\n",
+ " sleep_time = 0.1\n",
+ "\n",
+ " with open(last_backup_timestamps_path, \"w\") as f:\n",
+ " for filepath, timestamp in last_backup_timestamps.items():\n",
+ " f.write(f\"{filepath}:{timestamp}\\n\")\n",
+ "\n",
+ " time.sleep(sleep_time)\n",
+ "\n",
+ " except Exception as e:\n",
+ " print(f\"An error occurred: {str(e)}\")\n",
+ "\n",
+ "\n",
+ "if autobackups:\n",
+ " autobackups = False\n",
+ " print(\"Autobackup Disabled\")\n",
+ "else:\n",
+ " autobackups = True\n",
+ " print(\"Autobackup Enabled\")\n",
+ "\n",
+ "total_epoch = 800 # @param {type:\"integer\"}\n",
+ "batch_size = 15 # @param {type:\"slider\", min:1, max:25, step:0}\n",
+ "gpu = 0\n",
+ "sr = int(sample_rate.rstrip(\"k\")) * 1000\n",
+ "pitch_guidance = True # @param{type:\"boolean\"}\n",
+ "auto_backups = True # @param{type:\"boolean\"}\n",
+ "pretrained = True # @param{type:\"boolean\"}\n",
+ "sync_graph = False # @param{type:\"boolean\"}\n",
+ "tensorboard = True # @param{type:\"boolean\"}\n",
+ "# @markdown ### ➡️ Choose how many epochs your model will be stored\n",
+ "save_every_epoch = 10 # @param {type:\"slider\", min:1, max:100, step:0}\n",
+ "save_only_latest = False # @param{type:\"boolean\"}\n",
+ "save_every_weights = False # @param{type:\"boolean\"}\n",
+ "overtraining_detector = False # @param{type:\"boolean\"}\n",
+ "overtraining_threshold = 50 # @param {type:\"slider\", min:1, max:100, step:0}\n",
+ "# @markdown ### ❓ Optional\n",
+ "# @markdown In case you select custom pretrained, you will have to download the pretraineds and enter the path of the pretraineds.\n",
+ "custom_pretrained = False # @param{type:\"boolean\"}\n",
+ "g_pretrained_path = \"/content/Applio/rvc/pretraineds/pretraineds_custom/G48k.pth\" # @param {type:\"string\"}\n",
+ "d_pretrained_path = \"/content/Applio/rvc/pretraineds/pretraineds_custom/D48k.pth\" # @param {type:\"string\"}\n",
+ "\n",
+ "if \"pretrained\" not in globals():\n",
+ " pretrained = True\n",
+ "\n",
+ "if \"custom_pretrained\" not in globals():\n",
+ " custom_pretrained = False\n",
+ "\n",
+ "if \"g_pretrained_path\" not in globals():\n",
+ " g_pretrained_path = \"Custom Path\"\n",
+ "\n",
+ "if \"d_pretrained_path\" not in globals():\n",
+ " d_pretrained_path = \"Custom Path\"\n",
+ "\n",
+ "\n",
+ "def start_train():\n",
+ " if tensorboard == True:\n",
+ " %load_ext tensorboard\n",
+ " %tensorboard --logdir /content/Applio/logs/\n",
+ " !python core.py train --model_name \"{model_name}\" --rvc_version \"{rvc_version}\" --save_every_epoch \"{save_every_epoch}\" --save_only_latest \"{save_only_latest}\" --save_every_weights \"{save_every_weights}\" --total_epoch \"{total_epoch}\" --sampling_rate \"{sr}\" --batch_size \"{batch_size}\" --gpu \"{gpu}\" --pitch_guidance \"{pitch_guidance}\" --pretrained \"{pretrained}\" --custom_pretrained \"{custom_pretrained}\" --g_pretrained_path \"{g_pretrained_path}\" --d_pretrained_path \"{d_pretrained_path}\" --overtraining_detector \"{overtraining_detector}\" --overtraining_threshold \"{overtraining_threshold}\" --sync_graph \"{sync_graph}\"\n",
+ "\n",
+ "\n",
+ "server_thread = threading.Thread(target=start_train)\n",
+ "server_thread.start()\n",
+ "\n",
+ "if auto_backups:\n",
+ " backup_files()\n",
+ "else:\n",
+ " while True:\n",
+ " time.sleep(10)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "bHLs5AT4Q1ck"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Generate index file\n",
+ "!python core.py index --model_name \"{model_name}\" --rvc_version \"{rvc_version}\""
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "X_eU_SoiHIQg"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Save model\n",
+ "# @markdown Enter the name of the model and the steps. You can find it in your `/content/Applio/logs` folder.\n",
+ "%cd /content\n",
+ "import shutil, os\n",
+ "\n",
+ "model_name = \"Darwin\" # @param {type:\"string\"}\n",
+ "model_epoch = 800 # @param {type:\"integer\"}\n",
+ "save_big_file = False # @param {type:\"boolean\"}\n",
+ "\n",
+ "if os.path.exists(\"/content/zips\"):\n",
+ " shutil.rmtree(\"/content/zips\")\n",
+ "print(\"Removed zips.\")\n",
+ "!mkdir -p /content/zips/{model_name}/\n",
+ "print(\"Created zips.\")\n",
+ "if f\"{model_name}.pth\" not in os.listdir(f\"/content/Applio/weights\"):\n",
+ " print(\"There is no weight file with that name\")\n",
+ "if not save_big_file:\n",
+ " !cp /content/Applio/logs/{model_name}/added_*.index /content/zips/{model_name}/\n",
+ " !cp /content/Applio/logs/{model_name}/total_*.npy /content/zips/{model_name}/\n",
+ " !cp /content/Applio/weights/{model_name}.pth /content/zips/{model_name}/{model_name}{model_epoch}.pth\n",
+ " %cd /content/zips\n",
+ " !zip -r {model_name}.zip {model_name}\n",
+ "if save_big_file:\n",
+ " %cd /content/Applio\n",
+ " latest_steps = -1\n",
+ " logs_folder = \"./logs/\" + model_name\n",
+ " for filename in os.listdir(logs_folder):\n",
+ " if filename.startswith(\"G_\") and filename.endswith(\".pth\"):\n",
+ " steps = int(filename.split(\"_\")[1].split(\".\")[0])\n",
+ " if steps > latest_steps:\n",
+ " latest_steps = steps\n",
+ " MODELZIP = model_name + \".zip\"\n",
+ " !mkdir -p /content/zips\n",
+ " ZIPFILEPATH = os.path.join(\"/content/zips\", MODELZIP)\n",
+ " for filename in os.listdir(logs_folder):\n",
+ " if \"G_\" in filename or \"D_\" in filename:\n",
+ " if str(latest_steps) in filename:\n",
+ " !zip -r {ZIPFILEPATH} {os.path.join(logs_folder, filename)}\n",
+ " else:\n",
+ " !zip -r {ZIPFILEPATH} {os.path.join(logs_folder, filename)}\n",
+ " for filename in os.listdir(\"./weights\"):\n",
+ " if MODELNAME in filename:\n",
+ " !zip -r {ZIPFILEPATH} {os.path.join('./weights/', filename)}\n",
+ "\n",
+ "!mkdir -p /content/drive/MyDrive/RVC_Backup/\n",
+ "shutil.move(\n",
+ " f\"/content/zips/{model_name}.zip\",\n",
+ " f\"/content/drive/MyDrive/RVC_Backup/{model_name}.zip\",\n",
+ ")\n",
+ "%cd /content\n",
+ "shutil.rmtree(\"/content/zips\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "OaKoymXsyEYN"
+ },
+ "source": [
+ "# Resume-training"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "d3KgLAYnyHkP"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Load a Backup\n",
+ "from google.colab import drive\n",
+ "import os\n",
+ "import shutil\n",
+ "\n",
+ "# @markdown Put the exact name you put as your Model Name in Applio.\n",
+ "modelname = \"My-Project\" # @param {type:\"string\"}\n",
+ "source_path = \"/content/drive/MyDrive/RVC_Backup/\" + modelname\n",
+ "destination_path = \"/content/Applio/logs/\" + modelname\n",
+ "backup_timestamps_file = \"last_backup_timestamps.txt\"\n",
+ "if not os.path.exists(source_path):\n",
+ " print(\n",
+ " \"The model folder does not exist. Please verify the name is correct or check your Google Drive.\"\n",
+ " )\n",
+ "else:\n",
+ " time_ = os.path.join(\"/content/drive/MyDrive/RVC_Backup/\", backup_timestamps_file)\n",
+ " time__ = os.path.join(\"/content/Applio/logs/\", backup_timestamps_file)\n",
+ " if os.path.exists(time_):\n",
+ " shutil.copy(time_, time__)\n",
+ " shutil.copytree(source_path, destination_path)\n",
+ " print(\"Model backup loaded successfully.\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "sc9DzvRCyJ2d"
+ },
+ "outputs": [],
+ "source": [
+ "# @title Set training variables\n",
+ "# @markdown ### ➡️ Use the same as you did previously\n",
+ "model_name = \"Darwin\" # @param {type:\"string\"}\n",
+ "sample_rate = \"40k\" # @param [\"32k\", \"40k\", \"48k\"] {allow-input: false}\n",
+ "rvc_version = \"v2\" # @param [\"v2\", \"v1\"] {allow-input: false}\n",
+ "f0method = \"rmvpe\" # @param [\"pm\", \"dio\", \"crepe\", \"crepe-tiny\", \"harvest\", \"rmvpe\"] {allow-input: false}\n",
+ "hop_length = 128 # @param {type:\"slider\", min:1, max:512, step:0}\n",
+ "sr = int(sample_rate.rstrip(\"k\")) * 1000"
+ ]
+ }
+ ],
+ "metadata": {
+ "accelerator": "GPU",
+ "colab": {
+ "collapsed_sections": [
+ "ymMCTSD6m8qV"
+ ],
+ "provenance": [],
+ "toc_visible": true
+ },
+ "kernelspec": {
+ "display_name": "Python 3",
+ "name": "python3"
+ },
+ "language_info": {
+ "name": "python"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/assets/ICON.ico b/assets/ICON.ico
new file mode 100644
index 0000000000000000000000000000000000000000..340358a598d8a110c798431c8ca99bd580099b02
Binary files /dev/null and b/assets/ICON.ico differ
diff --git a/audios/.gitignore b/assets/audios/audio-others/.gitignore
similarity index 100%
rename from audios/.gitignore
rename to assets/audios/audio-others/.gitignore
diff --git a/assets/config.json b/assets/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..327e251d3731f8e4943e0d68bb9b98dcd21d9f14
--- /dev/null
+++ b/assets/config.json
@@ -0,0 +1,15 @@
+{
+ "theme": {
+ "file": "Applio.py",
+ "class": "Applio"
+ },
+ "plugins": [],
+ "discord_presence": true,
+ "lang": {
+ "override": false,
+ "selected_lang": "en_US"
+ },
+ "flask_server": false,
+ "version": "3.2.0",
+ "fake_gpu": false
+}
\ No newline at end of file
diff --git a/assets/delete_models.py b/assets/delete_models.py
new file mode 100644
index 0000000000000000000000000000000000000000..a4c047956a6afb6bc78f45be1b3323a8e5d8c50d
--- /dev/null
+++ b/assets/delete_models.py
@@ -0,0 +1,30 @@
+import sys
+import os
+import time
+import threading
+import shutil
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+
+
+def infinite_loop():
+ while True:
+ try:
+ models_folder = os.path.join(now_dir, "logs")
+
+ for element in os.listdir(models_folder):
+ element_route = os.path.join(models_folder, element)
+ if os.path.isdir(element_route) and element != "mute":
+ shutil.rmtree(element_route)
+ elif os.path.isfile(element_route):
+ os.remove(element_route)
+
+ wait_time = 24 * 60 * 60 #
+ time.sleep(wait_time)
+ except:
+ pass
+
+def start_infinite_loop():
+ hilo_bucle = threading.Thread(target=infinite_loop)
+ hilo_bucle.start()
\ No newline at end of file
diff --git a/assets/discord_presence.py b/assets/discord_presence.py
new file mode 100644
index 0000000000000000000000000000000000000000..d210f1ff3b02fa98c3005dd2f4ec05fab52caad1
--- /dev/null
+++ b/assets/discord_presence.py
@@ -0,0 +1,49 @@
+from pypresence import Presence
+import datetime as dt
+import time
+
+
+class RichPresenceManager:
+ def __init__(self):
+ self.client_id = "1144714449563955302"
+ self.rpc = None
+ self.running = False
+
+ def start_presence(self):
+ if not self.running:
+ self.running = True
+ self.rpc = Presence(self.client_id)
+ try:
+ self.rpc.connect()
+ self.update_presence()
+ except KeyboardInterrupt as error:
+ print(error)
+ self.rpc = None
+ self.running = False
+ except Exception as e:
+ print(f"Error: Unable to connect to Rich Presence. {e}")
+ self.rpc = None
+ self.running = False
+
+ def update_presence(self):
+ if self.rpc:
+ self.rpc.update(
+ state="applio.org",
+ details="Open ecosystem for voice cloning",
+ buttons=[
+ {"label": "Home", "url": "https://applio.org"},
+ {"label": "Download", "url": "https://applio.org/download"},
+ ],
+ large_image="logo",
+ large_text="Experimenting with applio",
+ start=dt.datetime.now().timestamp(),
+ )
+
+ def stop_presence(self):
+ self.running = False
+ if self.rpc:
+ self.rpc.close()
+ self.rpc = None
+
+
+RPCManager = RichPresenceManager()
diff --git a/assets/flask/routes.py b/assets/flask/routes.py
new file mode 100644
index 0000000000000000000000000000000000000000..8983e23945a0d3f811df78cc4be0abc6df8511e8
--- /dev/null
+++ b/assets/flask/routes.py
@@ -0,0 +1,32 @@
+import os, sys
+import signal
+from flask import Flask, request, redirect
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+
+from core import run_download_script
+
+app = Flask(__name__)
+
+
+@app.route("/download/", methods=["GET"])
+def download(url):
+ file_path = run_download_script(url)
+ if file_path == "Model downloaded successfully.":
+ if "text/html" in request.headers.get("Accept", ""):
+ return redirect("https://applio.org/models/downloaded", code=302)
+ else:
+ return ""
+ else:
+ return "Error: Unable to download file", 500
+
+
+@app.route("/shutdown", methods=["POST"])
+def shutdown():
+ print("This Flask server is shutting down... Please close the window!")
+ os.kill(os.getpid(), signal.SIGTERM)
+
+
+if __name__ == "__main__":
+ app.run(host="localhost", port=8000)
diff --git a/assets/flask/server.py b/assets/flask/server.py
new file mode 100644
index 0000000000000000000000000000000000000000..dc8697045ae704a939e2825bac82a3a37095dac9
--- /dev/null
+++ b/assets/flask/server.py
@@ -0,0 +1,62 @@
+import os
+import socket
+import subprocess
+import time
+import requests
+import sys
+import json
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+config_file = os.path.join(now_dir, "assets", "config.json")
+env_path = os.path.join(now_dir, "env", "python.exe")
+
+host = "localhost"
+port = 8000
+
+sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+sock.settimeout(2)
+
+
+def start_flask():
+ try:
+ sock.connect((host, port))
+ print(
+ f"Something is listening on port {port}; Probably the Flask server is already running."
+ )
+ print("Trying to start it anyway")
+ sock.close()
+ requests.post("http://localhost:8000/shutdown")
+ time.sleep(3)
+ script_path = os.path.join(now_dir, "assets", "flask", "routes.py")
+ try:
+ subprocess.Popen(
+ [env_path, script_path], creationflags=subprocess.CREATE_NEW_CONSOLE
+ )
+ except Exception as e:
+ print(f"Failed to start the Flask server")
+ print(e)
+ except Exception as e:
+ sock.close()
+ script_path = os.path.join(now_dir, "assets", "flask", "routes.py")
+ try:
+ subprocess.Popen(
+ [env_path, script_path], creationflags=subprocess.CREATE_NEW_CONSOLE
+ )
+ except Exception as e:
+ print("Failed to start the Flask server")
+ print(e)
+
+
+def load_config_flask():
+ with open(config_file, "r") as file:
+ config = json.load(file)
+ return config["flask_server"]
+
+
+def save_config(value):
+ with open(config_file, "r", encoding="utf8") as file:
+ config = json.load(file)
+ config["flask_server"] = value
+ with open(config_file, "w", encoding="utf8") as file:
+ json.dump(config, file, indent=2)
diff --git a/assets/hubert/.gitignore b/assets/hubert/.gitignore
deleted file mode 100644
index d6b7ef32c8478a48c3994dcadc86837f4371184d..0000000000000000000000000000000000000000
--- a/assets/hubert/.gitignore
+++ /dev/null
@@ -1,2 +0,0 @@
-*
-!.gitignore
diff --git a/assets/i18n/i18n.py b/assets/i18n/i18n.py
new file mode 100644
index 0000000000000000000000000000000000000000..c980c5b24f7338aa869217e089365b39e8086af8
--- /dev/null
+++ b/assets/i18n/i18n.py
@@ -0,0 +1,52 @@
+import os, sys
+import json
+from pathlib import Path
+from locale import getdefaultlocale
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+
+
+class I18nAuto:
+ LANGUAGE_PATH = os.path.join(now_dir, "assets", "i18n", "languages")
+
+ def __init__(self, language=None):
+ with open(
+ os.path.join(now_dir, "assets", "config.json"), "r", encoding="utf8"
+ ) as file:
+ config = json.load(file)
+ override = config["lang"]["override"]
+ lang_prefix = config["lang"]["selected_lang"]
+
+ self.language = lang_prefix
+
+ if override == False:
+ language = language or getdefaultlocale()[0]
+ lang_prefix = language[:2] if language is not None else "en"
+ available_languages = self._get_available_languages()
+ matching_languages = [
+ lang for lang in available_languages if lang.startswith(lang_prefix)
+ ]
+ self.language = matching_languages[0] if matching_languages else "en_US"
+
+ self.language_map = self._load_language_list()
+
+ def _load_language_list(self):
+ try:
+ file_path = Path(self.LANGUAGE_PATH) / f"{self.language}.json"
+ with open(file_path, "r", encoding="utf-8") as file:
+ return json.load(file)
+ except FileNotFoundError:
+ raise FileNotFoundError(
+ f"Failed to load language file for {self.language}. Check if the correct .json file exists."
+ )
+
+ def _get_available_languages(self):
+ language_files = [path.stem for path in Path(self.LANGUAGE_PATH).glob("*.json")]
+ return language_files
+
+ def _language_exists(self, language):
+ return (Path(self.LANGUAGE_PATH) / f"{language}.json").exists()
+
+ def __call__(self, key):
+ return self.language_map.get(key, key)
diff --git a/assets/i18n/languages/ar_AR.json b/assets/i18n/languages/ar_AR.json
new file mode 100644
index 0000000000000000000000000000000000000000..af91a9c943d6daa25c8928ed405524b556174506
--- /dev/null
+++ b/assets/i18n/languages/ar_AR.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "أداة استنساخ الصوت النهائية ، محسنة بدقة للحصول على قوة لا مثيل لها ، ونمطية ، وتجربة سهلة الاستخدام.",
+ "This section contains some extra utilities that often may be in experimental phases.": "يحتوي هذا القسم على بعض الأدوات المساعدة الإضافية التي قد تكون غالبا في المراحل التجريبية.",
+ "Output Information": "معلومات الإخراج",
+ "The output information will be displayed here.": "سيتم عرض معلومات الإخراج هنا.",
+ "Inference": "استدلال",
+ "Train": "قطار",
+ "Extra": "اضافيه",
+ "Merge Audios": "دمج الصوتيات",
+ "Processing": "تجهيز",
+ "Audio Analyzer": "محلل الصوت",
+ "Model Information": "معلومات النموذج",
+ "Plugins": "الإضافات",
+ "Download": "تحميل",
+ "Report a Bug": "الإبلاغ عن خطأ",
+ "Settings": "اعدادات",
+ "Preprocess": "المعالجة المسبقة",
+ "Model Name": "اسم الموديل",
+ "Name of the new model.": "اسم النموذج الجديد.",
+ "Enter model name": "أدخل اسم الطراز",
+ "Dataset Path": "مسار مجموعة البيانات",
+ "Path to the dataset folder.": "المسار إلى مجلد مجموعة البيانات.",
+ "Refresh Datasets": "تحديث مجموعات البيانات",
+ "Dataset Creator": "منشئ مجموعة البيانات",
+ "Dataset Name": "اسم مجموعة البيانات",
+ "Name of the new dataset.": "اسم مجموعة البيانات الجديدة.",
+ "Enter dataset name": "أدخل اسم مجموعة البيانات",
+ "Upload Audio Dataset": "تحميل مجموعة بيانات صوتية",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "تمت إضافة الملف الصوتي بنجاح إلى مجموعة البيانات. الرجاء النقر فوق زر المعالجة المسبقة.",
+ "Enter dataset path": "إدخال مسار مجموعة البيانات",
+ "Sampling Rate": "معدل أخذ العينات",
+ "The sampling rate of the audio files.": "معدل أخذ العينات من الملفات الصوتية.",
+ "RVC Version": "نسخة RVC",
+ "The RVC version of the model.": "نسخة RVC من النموذج.",
+ "Preprocess Dataset": "مجموعة بيانات ما قبل المعالجة",
+ "Extract": "استخرج",
+ "Hop Length": "طول القفزة",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "يشير إلى المدة التي يستغرقها النظام للانتقال إلى تغيير كبير في درجة الصوت. تتطلب أطوال القفزات الأصغر مزيدا من الوقت للاستدلال ولكنها تميل إلى تحقيق دقة أعلى في درجة الصوت.",
+ "Batch Size": "حجم الدفعة",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "ينصح بمحاذاته مع VRAM المتاح لوحدة معالجة الرسومات الخاصة بك. يوفر الإعداد 4 دقة محسنة ولكن معالجة أبطأ ، بينما يوفر 8 نتائج أسرع وقياسية.",
+ "Save Every Epoch": "حفظ كل حقبة",
+ "Determine at how many epochs the model will saved at.": "حدد عدد الفترات التي سيتم حفظ النموذج فيها.",
+ "Total Epoch": "إجمالي العصر",
+ "Specifies the overall quantity of epochs for the model training process.": "يحدد الكمية الإجمالية للعهود لعملية التدريب النموذجية.",
+ "Pretrained": "التدريب المسبق",
+ "Save Only Latest": "حفظ الأحدث فقط",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "سيؤدي تمكين هذا الإعداد إلى حفظ ملفات G و D لأحدث إصداراتها فقط ، مما يوفر مساحة التخزين بشكل فعال.",
+ "Save Every Weights": "حفظ كل الأوزان",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "يمكنك هذا الإعداد من حفظ أوزان النموذج في نهاية كل حقبة.",
+ "Custom Pretrained": "تدريب مسبق مخصص",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "يمكن أن يؤدي استخدام النماذج المخصصة المدربة مسبقا إلى نتائج فائقة ، حيث أن اختيار النماذج الأكثر ملاءمة للاختبار المسبق والمصممة خصيصا لحالة الاستخدام المحددة يمكن أن يعزز الأداء بشكل كبير.",
+ "Upload Pretrained Model": "تحميل نموذج تم تدريبه مسبقا",
+ "Refresh Custom Pretraineds": "تحديث التدريبات المسبقة المخصصة",
+ "Pretrained Custom Settings": "الإعدادات المخصصة المدربة مسبقا",
+ "The file you dropped is not a valid pretrained file. Please try again.": "الملف الذي أسقطته ليس ملفا صالحا تم تدريبه مسبقا. يرجى المحاولة مرة أخرى.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "انقر فوق زر التحديث لرؤية الملف الذي تم اختباره مسبقا في القائمة المنسدلة.",
+ "Pretrained G Path": "مخصص مسبقا G",
+ "Pretrained D Path": "مخصص مسبق التدريب D",
+ "GPU Settings": "إعدادات وحدة معالجة الرسومات",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "يضبط إعدادات GPU المتقدمة ، الموصى بها للمستخدمين الذين لديهم بنية GPU أفضل.",
+ "GPU Custom Settings": "الإعدادات المخصصة لوحدة معالجة الرسومات",
+ "GPU Number": "رقم وحدة معالجة الرسومات",
+ "0 to ∞ separated by -": "0 إلى ∞ مفصولة ب -",
+ "GPU Information": "معلومات وحدة معالجة الرسومات",
+ "Pitch Guidance": "توجيه الملعب",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "من خلال استخدام توجيه درجة الصوت ، يصبح من الممكن عكس نغمة الصوت الأصلي ، بما في ذلك طبقة الصوت. هذه الميزة ذات قيمة خاصة للغناء والسيناريوهات الأخرى حيث يكون الحفاظ على اللحن الأصلي أو نمط طبقة الصوت أمرا ضروريا.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "استخدم النماذج المدربة مسبقا عند تدريب النماذج الخاصة بك. هذا النهج يقلل من مدة التدريب ويعزز الجودة الشاملة.",
+ "Extract Features": "استخراج الميزات",
+ "Start Training": "ابدأ التدريب",
+ "Generate Index": "إنشاء فهرس",
+ "Voice Model": "نموذج الصوت",
+ "Select the voice model to use for the conversion.": "حدد نموذج الصوت لاستخدامه في التحويل.",
+ "Index File": "ملف الفهرس",
+ "Select the index file to use for the conversion.": "حدد ملف الفهرس لاستخدامه للتحويل.",
+ "Refresh": "تحديث",
+ "Unload Voice": "تفريغ الصوت",
+ "Single": "واحد",
+ "Upload Audio": "تحميل الصوت",
+ "Select Audio": "حدد الصوت",
+ "Select the audio to convert.": "حدد الصوت المراد تحويله.",
+ "Advanced Settings": "الإعدادات المتقدمة",
+ "Clear Outputs (Deletes all audios in assets/audios)": "مخرجات واضحة (يحذف جميع الصوتيات في الأصول / الصوتيات)",
+ "Custom Output Path": "مسار الإخراج المخصص",
+ "Output Path": "مسار الإخراج",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "المسار الذي سيتم فيه حفظ الصوت الناتج ، افتراضيا في الأصول / الصوتيات / output.wav",
+ "Split Audio": "تقسيم الصوت",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "قسم الصوت إلى أجزاء للاستدلال للحصول على نتائج أفضل في بعض الحالات.",
+ "Autotune": "الضبط التلقائي",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "قم بتطبيق ضبط تلقائي ناعم على استنتاجاتك ، موصى به لغناء التحويلات.",
+ "Clean Audio": "صوت نظيف",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "قم بتنظيف إخراج الصوت باستخدام خوارزميات اكتشاف الضوضاء ، الموصى بها للتحدث الصوتيات.",
+ "Clean Strength": "قوة نظيفة",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "اضبط مستوى التنظيف على الصوت الذي تريده ، وكلما قمت بزيادته كلما تم تنظيفه ، ولكن من الممكن أن يكون الصوت أكثر ضغطا.",
+ "Pitch": "زفت",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "اضبط درجة الصوت ، وكلما زادت القيمة ، زادت درجة الصوت.",
+ "Filter Radius": "نصف قطر المرشح",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "إذا كان العدد أكبر من أو يساوي ثلاثة ، فإن استخدام الترشيح المتوسط على نتائج النغمة التي تم جمعها لديه القدرة على تقليل التنفس.",
+ "Search Feature Ratio": "نسبة ميزة البحث",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "التأثير الذي يمارسه ملف الفهرس ؛ قيمة أعلى يتوافق مع تأثير أكبر. ومع ذلك ، يمكن أن يساعد اختيار القيم الأقل في التخفيف من العناصر الموجودة في الصوت.",
+ "Volume Envelope": "مغلف الحجم",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "استبدل أو امزج مع مغلف حجم المخرجات. كلما اقتربت النسبة من 1 ، زاد استخدام مغلف الإخراج.",
+ "Protect Voiceless Consonants": "حماية الحروف الساكنة التي لا صوت لها",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "حماية الحروف الساكنة المميزة وأصوات التنفس لمنع التمزق الصوتي الكهربائي والتحف الأخرى. يوفر سحب المعلمة إلى قيمتها القصوى البالغة 0.5 حماية شاملة. ومع ذلك ، قد يؤدي تقليل هذه القيمة إلى تقليل مدى الحماية مع احتمال التخفيف من تأثير الفهرسة.",
+ "Pitch extraction algorithm": "خوارزمية استخراج الملعب",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "خوارزمية استخراج الملعب لاستخدامها في تحويل الصوت. الخوارزمية الافتراضية هي rmvpe ، والتي يوصى بها في معظم الحالات.",
+ "Convert": "حول",
+ "Export Audio": "تصدير الصوت",
+ "Batch": "الدفعه",
+ "Input Folder": "مجلد الإدخال",
+ "Select the folder containing the audios to convert.": "حدد المجلد الذي يحتوي على الصوتيات المراد تحويلها.",
+ "Enter input path": "أدخل مسار الإدخال",
+ "Output Folder": "مجلد الإخراج",
+ "Select the folder where the output audios will be saved.": "حدد المجلد حيث سيتم حفظ صوتيات الإخراج.",
+ "Enter output path": "أدخل مسار الإخراج",
+ "Get information about the audio": "الحصول على معلومات حول الصوت",
+ "Information about the audio file": "معلومات حول الملف الصوتي",
+ "Waiting for information...": "في انتظار المعلومات...",
+ "## Voice Blender": "## خلاط الصوت",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "حدد نموذجين صوتيين ، وقم بتعيين نسبة المزج التي تريدها ، وامزجهما في صوت جديد تماما.",
+ "Voice Blender": "خلاط الصوت",
+ "Drag and drop your model here": "قم بسحب وإسقاط النموذج الخاص بك هنا",
+ "You can also use a custom path.": "يمكنك أيضا استخدام مسار مخصص.",
+ "Blend Ratio": "نسبة المزج",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "سيؤدي ضبط الموضع أكثر نحو جانب أو آخر إلى جعل النموذج أكثر تشابها مع الأول أو الثاني.",
+ "Fusion": "اندماج",
+ "Path to Model": "الطريق إلى النموذج",
+ "Enter path to model": "أدخل المسار إلى النموذج",
+ "Model information to be placed": "معلومات النموذج المراد وضعها",
+ "Inroduce the model information": "Inroduce معلومات النموذج",
+ "The information to be placed in the model (You can leave it blank or put anything).": "المعلومات المراد وضعها في النموذج (يمكنك تركها فارغة أو وضع أي شيء).",
+ "View model information": "عرض معلومات النموذج",
+ "Introduce the model pth path": "تقديم نموذج مسار pth",
+ "View": "منظر",
+ "Model extraction": "استخراج النموذج",
+ "Model conversion": "تحويل النموذج",
+ "Pth file": "ملف Pth",
+ "Output of the pth file": "إخراج ملف pth",
+ "# How to Report an Issue on GitHub": "# كيفية الإبلاغ عن مشكلة على GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. انقر فوق الزر \"شاشة التسجيل\" أدناه لبدء تسجيل المشكلة التي تواجهها.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. بمجرد الانتهاء من تسجيل المشكلة ، انقر فوق الزر \"إيقاف التسجيل\" (نفس الزر ، لكن التسمية تتغير اعتمادا على ما إذا كنت تقوم بالتسجيل بنشاط أم لا).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. انتقل إلى [مشكلات GitHub] (https://github.com/IAHispano/Applio/issues) وانقر على زر \"إصدار جديد\".",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. أكمل نموذج المشكلة المقدم ، مع التأكد من تضمين التفاصيل حسب الحاجة ، واستخدم قسم الأصول لتحميل الملف المسجل من الخطوة السابقة.",
+ "Record Screen": "شاشة التسجيل",
+ "Record": "سجل",
+ "Stop Recording": "إيقاف التسجيل",
+ "Introduce the model .pth path": "تقديم نموذج مسار .pth",
+ "See Model Information": "انظر معلومات النموذج",
+ "## Download Model": "## تحميل الموديل",
+ "Model Link": "رابط النموذج",
+ "Introduce the model link": "تقديم رابط النموذج",
+ "Download Model": "تحميل الموديل",
+ "## Drop files": "## إسقاط الملفات",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "اسحب ملف .pth وملف .index إلى هذه المساحة. اسحب أحدهما ثم الآخر.",
+ "TTS Voices": "أصوات تحويل النص إلى كلام",
+ "Select the TTS voice to use for the conversion.": "حدد صوت TTS لاستخدامه في التحويل.",
+ "Text to Synthesize": "النص المراد توليفه",
+ "Enter the text to synthesize.": "أدخل النص المراد توليفه.",
+ "Or you can upload a .txt file": "أو يمكنك تحميل ملف .txt",
+ "Enter text to synthesize": "أدخل نصا لتوليفه",
+ "Output Path for TTS Audio": "مسار الإخراج لصوت TTS",
+ "Output Path for RVC Audio": "مسار الإخراج لصوت RVC",
+ "Enable Applio integration with Discord presence": "تمكين تكامل Applio مع وجود Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "سيتم تنشيط إمكانية عرض نشاط Applio الحالي في Discord.",
+ "Enable Applio integration with applio.org/models using flask": "تمكين تكامل Applio مع applio.org/models باستخدام القارورة",
+ "It will activate the possibility of downloading models with a click from the website.": "سيتم تنشيط إمكانية تنزيل النماذج بنقرة واحدة من الموقع.",
+ "Theme": "موضوع",
+ "Select the theme you want to use. (Requires restarting Applio)": "حدد السمة التي تريد استخدامها. (يتطلب إعادة تشغيل Applio)",
+ "Language": "اللغة",
+ "Select the language you want to use. (Requires restarting Applio)": "حدد اللغة التي تريد استخدامها. (يتطلب إعادة تشغيل Applio)",
+ "Plugin Installer": "مثبت البرنامج المساعد",
+ "Drag your plugin.zip to install it": "اسحب plugin.zip لتثبيته",
+ "Version Checker": "مدقق الإصدار",
+ "Check which version of Applio is the latest to see if you need to update.": "تحقق من إصدار Applio هو الأحدث لمعرفة ما إذا كنت بحاجة إلى التحديث.",
+ "Check for updates": "التحقق من وجود تحديثات"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/bn_BN.json b/assets/i18n/languages/bn_BN.json
new file mode 100644
index 0000000000000000000000000000000000000000..7a7f5462268c0797b64ae7134ab719b3bfc37a00
--- /dev/null
+++ b/assets/i18n/languages/bn_BN.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "আলটিমেট ভয়েস ক্লোনিং টুল, অতুলনীয় শক্তি, মডুলারিটি এবং ব্যবহারকারী-বান্ধব অভিজ্ঞতার জন্য নিখুঁতভাবে অপ্টিমাইজ করা।",
+ "This section contains some extra utilities that often may be in experimental phases.": "এই বিভাগে কিছু অতিরিক্ত ইউটিলিটি রয়েছে যা প্রায়শই পরীক্ষামূলক পর্যায়ে থাকতে পারে।",
+ "Output Information": "আউটপুট তথ্য",
+ "The output information will be displayed here.": "আউটপুট তথ্য এখানে প্রদর্শিত হবে।",
+ "Inference": "অনুমান",
+ "Train": "ট্রেন",
+ "Extra": "অতিরিক্ত",
+ "Merge Audios": "অডিওগুলি মার্জ করুন",
+ "Processing": "প্রক্রিয়াকরণ",
+ "Audio Analyzer": "অডিও বিশ্লেষক",
+ "Model Information": "মডেল তথ্য",
+ "Plugins": "প্লাগইন",
+ "Download": "ডাউনলোড",
+ "Report a Bug": "একটি বাগ রিপোর্ট করুন",
+ "Settings": "সেটিংস",
+ "Preprocess": "প্রিপ্রসেস",
+ "Model Name": "মডেলের নাম",
+ "Name of the new model.": "নতুন মডেলের নাম",
+ "Enter model name": "মডেলের নাম লিখুন",
+ "Dataset Path": "ডেটাসেট পাথ",
+ "Path to the dataset folder.": "ডেটাসেট ফোল্ডারে পাথ।",
+ "Refresh Datasets": "ডেটাসেট রিফ্রেশ করুন",
+ "Dataset Creator": "ডেটাসেট স্রষ্টা",
+ "Dataset Name": "ডেটাসেটের নাম",
+ "Name of the new dataset.": "নতুন ডেটাসেটের নাম",
+ "Enter dataset name": "তথ্যসেটের নাম লিখুন",
+ "Upload Audio Dataset": "অডিও ডেটাসেট আপলোড করুন",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "অডিও ফাইলটি সফলভাবে ডেটাসেটে যুক্ত করা হয়েছে। অনুগ্রহ করে প্রিপ্রসেস বাটনে ক্লিক করুন।",
+ "Enter dataset path": "ডেটাসেটের পথ লিখুন",
+ "Sampling Rate": "নমুনা হার",
+ "The sampling rate of the audio files.": "অডিও ফাইলের নমুনা হার।",
+ "RVC Version": "আরভিসি সংস্করণ",
+ "The RVC version of the model.": "মডেলটির আরভিসি সংস্করণ।",
+ "Preprocess Dataset": "প্রিপ্রসেস ডেটাসেট",
+ "Extract": "নিষ্কাশন",
+ "Hop Length": "হপ দৈর্ঘ্য",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "সিস্টেমটি একটি উল্লেখযোগ্য পিচ পরিবর্তনে রূপান্তরিত হতে যে সময়কাল নেয় তা বোঝায়। ছোট হপ দৈর্ঘ্যের জন্য অনুমানের জন্য আরও সময় প্রয়োজন তবে উচ্চতর পিচ নির্ভুলতা অর্জন করে।",
+ "Batch Size": "ব্যাচের আকার",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "এটি আপনার জিপিইউর উপলব্ধ ভিআরএএমের সাথে সারিবদ্ধ করার পরামর্শ দেওয়া হচ্ছে। 4 এর একটি সেটিং উন্নত নির্ভুলতা সরবরাহ করে তবে ধীর প্রক্রিয়াজাতকরণ, যখন 8 দ্রুত এবং মানক ফলাফল সরবরাহ করে।",
+ "Save Every Epoch": "প্রতিটি যুগ সংরক্ষণ করুন",
+ "Determine at how many epochs the model will saved at.": "মডেলটি কতগুলি যুগে সংরক্ষণ করবে তা নির্ধারণ করুন।",
+ "Total Epoch": "মোট যুগ",
+ "Specifies the overall quantity of epochs for the model training process.": "মডেল প্রশিক্ষণ প্রক্রিয়ার জন্য যুগের সামগ্রিক পরিমাণ উল্লেখ করে।",
+ "Pretrained": "পূর্বনির্ধারিত",
+ "Save Only Latest": "শুধুমাত্র সর্বশেষ সংরক্ষণ করুন",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "এই সেটিংটি সক্ষম করার ফলে জি এবং ডি ফাইলগুলি কেবলমাত্র তাদের সাম্প্রতিকতম সংস্করণগুলি সংরক্ষণ করবে, কার্যকরভাবে স্টোরেজ স্পেস সংরক্ষণ করবে।",
+ "Save Every Weights": "প্রতিটি ওজন সংরক্ষণ করুন",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "এই সেটিংটি আপনাকে প্রতিটি যুগের শেষে মডেলের ওজন সংরক্ষণ করতে সক্ষম করে।",
+ "Custom Pretrained": "কাস্টম প্রিট্রেইনড",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "কাস্টম প্রিট্রেনড মডেলগুলি ব্যবহার করা উচ্চতর ফলাফলের দিকে পরিচালিত করতে পারে, কারণ নির্দিষ্ট ব্যবহারের ক্ষেত্রে উপযুক্ত প্রিট্রেনড মডেলগুলি নির্বাচন করা কর্মক্ষমতা উল্লেখযোগ্যভাবে বাড়িয়ে তুলতে পারে।",
+ "Upload Pretrained Model": "প্রিট্রেনড মডেল আপলোড করুন",
+ "Refresh Custom Pretraineds": "কাস্টম প্রিট্রেনেডগুলি রিফ্রেশ করুন",
+ "Pretrained Custom Settings": "পূর্বনির্ধারিত কাস্টম সেটিংস",
+ "The file you dropped is not a valid pretrained file. Please try again.": "আপনার ফেলে দেওয়া ফাইলটি একটি বৈধ পূর্বপ্রশিক্ষিত ফাইল নয়. অনুগ্রহ করে আবার চেষ্টা করুন।",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "ড্রপডাউন মেনুতে প্রিট্রেনড ফাইলটি দেখতে রিফ্রেশ বোতামটি ক্লিক করুন।",
+ "Pretrained G Path": "কাস্টম প্রিট্রেনড জি",
+ "Pretrained D Path": "কাস্টম প্রিট্রেনড ডি",
+ "GPU Settings": "জিপিইউ সেটিংস",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "উন্নত GPU সেটিংস সেট করে, আরও ভাল GPU আর্কিটেকচার সহ ব্যবহারকারীদের জন্য প্রস্তাবিত।",
+ "GPU Custom Settings": "GPU কাস্টম সেটিংস",
+ "GPU Number": "জিপিইউ নম্বর",
+ "0 to ∞ separated by -": "0 থেকে ∞ দ্বারা পৃথক করা হয় -",
+ "GPU Information": "জিপিইউ তথ্য",
+ "Pitch Guidance": "পিচ গাইডেন্স",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "পিচ গাইডেন্স নিয়োগ করে, এর পিচ সহ মূল ভয়েসের স্বরভঙ্গিটি মিরর করা সম্ভব হয়। এই বৈশিষ্ট্যটি গাওয়া এবং অন্যান্য পরিস্থিতিতে বিশেষত মূল্যবান যেখানে মূল সুর বা পিচ প্যাটার্ন সংরক্ষণ করা অপরিহার্য।",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "আপনার নিজের প্রশিক্ষণের সময় প্রিপ্রশিক্ষিত মডেলগুলি ব্যবহার করুন। এই পদ্ধতির প্রশিক্ষণের সময়কাল হ্রাস করে এবং সামগ্রিক মান বাড়ায়।",
+ "Extract Features": "এক্সট্রাক্ট বৈশিষ্ট্য",
+ "Start Training": "প্রশিক্ষণ শুরু করুন",
+ "Generate Index": "সূচী তৈরি করুন",
+ "Voice Model": "ভয়েস মডেল",
+ "Select the voice model to use for the conversion.": "রূপান্তরটির জন্য ব্যবহার করতে ভয়েস মডেলটি নির্বাচন করুন।",
+ "Index File": "সূচী ফাইল",
+ "Select the index file to use for the conversion.": "রূপান্তরটির জন্য ব্যবহার করতে সূচী ফাইলটি নির্বাচন করুন।",
+ "Refresh": "সতেজ",
+ "Unload Voice": "ভয়েস আনলোড করুন",
+ "Single": "একক",
+ "Upload Audio": "অডিও আপলোড করুন",
+ "Select Audio": "অডিও নির্বাচন করুন",
+ "Select the audio to convert.": "রূপান্তর করতে অডিও নির্বাচন করুন।",
+ "Advanced Settings": "উন্নত সেটিংস",
+ "Clear Outputs (Deletes all audios in assets/audios)": "আউটপুট সাফ করুন (সম্পদ / অডিওতে সমস্ত অডিও মুছে ফেলে)",
+ "Custom Output Path": "কাস্টম আউটপুট পাথ",
+ "Output Path": "আউটপুট পাথ",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "পাথ যেখানে আউটপুট অডিও সংরক্ষণ করা হবে, সম্পদ / অডিও / output.wav ডিফল্টরূপে",
+ "Split Audio": "অডিও বিভক্ত করুন",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "কিছু ক্ষেত্রে আরও ভাল ফলাফল পেতে অনুমানের জন্য অডিওটিকে খণ্ডগুলিতে বিভক্ত করুন।",
+ "Autotune": "অটোটিউন",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "আপনার অনুমানগুলিতে একটি নরম অটোটিউন প্রয়োগ করুন, রূপান্তরগুলি গাওয়ার জন্য প্রস্তাবিত।",
+ "Clean Audio": "পরিষ্কার অডিও",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "অডিও বলার জন্য প্রস্তাবিত কোলাহল শনাক্তকরণ অ্যালগরিদমগুলি ব্যবহার করে আপনার অডিও আউটপুট পরিষ্কার করুন।",
+ "Clean Strength": "পরিষ্কার শক্তি",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "আপনি যে অডিওটি চান তাতে ক্লিন-আপ স্তরটি সেট করুন, আপনি এটি যত বাড়াবেন তত বেশি এটি পরিষ্কার হবে, তবে এটি সম্ভব যে অডিওটি আরও সংকুচিত হবে।",
+ "Pitch": "পিচ",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "অডিওর পিচ সেট করুন, মান যত বেশি, পিচ তত বেশি।",
+ "Filter Radius": "ফিল্টার ব্যাসার্ধ",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "যদি সংখ্যাটি তিনটির চেয়ে বেশি বা সমান হয় তবে সংগৃহীত স্বন ফলাফলগুলিতে মধ্যমা ফিল্টারিং নিয়োগ করা শ্বাসকষ্ট হ্রাস করার সম্ভাবনা রয়েছে।",
+ "Search Feature Ratio": "অনুসন্ধান বৈশিষ্ট্য অনুপাত",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "ইনডেক্স ফাইল দ্বারা প্রভাবিত; একটি উচ্চতর মান বৃহত্তর প্রভাবের সাথে মিলে যায়। তবে, নিম্ন মানগুলি বেছে নেওয়া অডিওতে উপস্থিত নিদর্শনগুলি প্রশমিত করতে সহায়তা করতে পারে।",
+ "Volume Envelope": "ভলিউম খাম",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "আউটপুটের ভলিউম খামের সাথে বিকল্প বা মিশ্রণ করুন। অনুপাতটি 1 এর কাছাকাছি হয়, তত বেশি আউটপুট খাম নিযুক্ত করা হয়।",
+ "Protect Voiceless Consonants": "কণ্ঠহীন ব্যঞ্জনবর্ণ রক্ষা করুন",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "ইলেক্ট্রো-অ্যাকোস্টিক ছিঁড়ে যাওয়া এবং অন্যান্য নিদর্শনগুলি রোধ করতে স্বতন্ত্র ব্যঞ্জনবর্ণ এবং শ্বাস প্রশ্বাসের শব্দগুলি রক্ষা করুন। প্যারামিটারটিকে তার সর্বোচ্চ মান 0.5 এ টানলে ব্যাপক সুরক্ষা সরবরাহ করে। যাইহোক, এই মান হ্রাস করা সম্ভাব্যভাবে সূচক প্রভাব প্রশমিত করার সময় সুরক্ষার পরিমাণ হ্রাস করতে পারে।",
+ "Pitch extraction algorithm": "পিচ নিষ্কাশন অ্যালগরিদম",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "অডিও রূপান্তর জন্য ব্যবহার করতে পিচ নিষ্কাশন অ্যালগরিদম. ডিফল্ট অ্যালগরিদমটি আরএমভিপিই, যা বেশিরভাগ ক্ষেত্রে প্রস্তাবিত।",
+ "Convert": "রূপান্তর",
+ "Export Audio": "অডিও রপ্তানি করুন",
+ "Batch": "ব্যাচ",
+ "Input Folder": "ইনপুট ফোল্ডার",
+ "Select the folder containing the audios to convert.": "রূপান্তর করতে অডিওযুক্ত ফোল্ডারটি নির্বাচন করুন।",
+ "Enter input path": "ইনপুট পথ লিখুন",
+ "Output Folder": "আউটপুট ফোল্ডার",
+ "Select the folder where the output audios will be saved.": "ফোল্ডারটি নির্বাচন করুন যেখানে আউটপুট অডিওগুলি সংরক্ষণ করা হবে।",
+ "Enter output path": "আউটপুট পথ লিখুন",
+ "Get information about the audio": "অডিও সম্পর্কে তথ্য পান",
+ "Information about the audio file": "অডিও ফাইল সম্পর্কে তথ্য",
+ "Waiting for information...": "তথ্যের অপেক্ষায়...",
+ "## Voice Blender": "## ভয়েস ব্লেন্ডার",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "দুটি ভয়েস মডেল নির্বাচন করুন, আপনার পছন্দসই মিশ্রণের শতাংশ সেট করুন এবং এগুলি সম্পূর্ণ নতুন ভয়েসে মিশ্রিত করুন।",
+ "Voice Blender": "ভয়েস ব্লেন্ডার",
+ "Drag and drop your model here": "আপনার মডেলটি এখানে টেনে এনে ছেড়ে দিন",
+ "You can also use a custom path.": "আপনি একটি কাস্টম পাথও ব্যবহার করতে পারেন।",
+ "Blend Ratio": "ব্লেন্ড রেশিও",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "একপাশে বা অন্যদিকে অবস্থানটি আরও সামঞ্জস্য করা মডেলটিকে প্রথম বা দ্বিতীয়টির সাথে আরও অনুরূপ করে তুলবে।",
+ "Fusion": "ফিউশন",
+ "Path to Model": "মডেলের পথ",
+ "Enter path to model": "মডেলের পথ লিখুন",
+ "Model information to be placed": "মডেল তথ্য স্থাপন করা হবে",
+ "Inroduce the model information": "মডেলের তথ্য ইনরোডিউস করুন",
+ "The information to be placed in the model (You can leave it blank or put anything).": "মডেলটিতে যে তথ্য রাখতে হবে (আপনি এটি ফাঁকা রেখে দিতে পারেন বা কিছু রাখতে পারেন)।",
+ "View model information": "মডেল তথ্য দেখুন",
+ "Introduce the model pth path": "মডেল পিটিএইচ পাথ পরিচয় করিয়ে দিন",
+ "View": "দর্শন",
+ "Model extraction": "মডেল নিষ্কাশন",
+ "Model conversion": "মডেল রূপান্তর",
+ "Pth file": "Pth ফাইল",
+ "Output of the pth file": "পিটিএইচ ফাইলের আউটপুট",
+ "# How to Report an Issue on GitHub": "# গিটহাবে একটি সমস্যা কিভাবে রিপোর্ট করবেন",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. আপনি যে সমস্যার সম্মুখীন হচ্ছেন তা রেকর্ড করা শুরু করতে নীচের 'রেকর্ড স্ক্রিন' বোতামে ক্লিক করুন।",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. একবার আপনি সমস্যাটি রেকর্ড করা শেষ করার পরে, 'রেকর্ডিং বন্ধ করুন' বোতামে ক্লিক করুন (একই বোতাম, তবে আপনি সক্রিয়ভাবে রেকর্ড করছেন কিনা তার উপর নির্ভর করে লেবেলটি পরিবর্তিত হয়)।",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "৩. [GitHub Issues](https://github.com/IAHispano/Applio/issues) এ যান এবং 'New Issue' বাটনে ক্লিক করুন।",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. প্রদত্ত ইস্যু টেমপ্লেটটি সম্পূর্ণ করুন, প্রয়োজন অনুসারে বিশদ অন্তর্ভুক্ত করা নিশ্চিত করুন এবং পূর্ববর্তী পদক্ষেপ থেকে রেকর্ড করা ফাইলটি আপলোড করতে সম্পদ বিভাগটি ব্যবহার করুন।",
+ "Record Screen": "রেকর্ড স্ক্রিন",
+ "Record": "রেকর্ড",
+ "Stop Recording": "রেকর্ডিং বন্ধ করুন",
+ "Introduce the model .pth path": "মডেল .pth পাথ পরিচয় করিয়ে দিন",
+ "See Model Information": "মডেল তথ্য দেখুন",
+ "## Download Model": "## মডেল ডাউনলোড করুন",
+ "Model Link": "মডেল লিংক",
+ "Introduce the model link": "মডেল লিঙ্কটি পরিচয় করিয়ে দিন",
+ "Download Model": "মডেল ডাউনলোড করুন",
+ "## Drop files": "## ফাইল ড্রপ করুন",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "আপনার .pth ফাইল এবং .index ফাইলটি এই স্পেসে টেনে আনুন। একটা টেনে আনুন, তারপর অন্যটা।",
+ "TTS Voices": "টিটিএস ভয়েসেস",
+ "Select the TTS voice to use for the conversion.": "রূপান্তরটির জন্য ব্যবহার করতে TTS ভয়েস নির্বাচন করুন।",
+ "Text to Synthesize": "সংশ্লেষণ করার জন্য পাঠ্য",
+ "Enter the text to synthesize.": "সংশ্লেষ করতে পাঠ্যটি প্রবেশ করান।",
+ "Or you can upload a .txt file": "অথবা আপনি একটি .txt ফাইল আপলোড করতে পারেন",
+ "Enter text to synthesize": "সংশ্লেষ করতে পাঠ্য লিখুন",
+ "Output Path for TTS Audio": "TTS অডিওর জন্য আউটপুট পাথ",
+ "Output Path for RVC Audio": "আরভিসি অডিওর জন্য আউটপুট পাথ",
+ "Enable Applio integration with Discord presence": "ডিসকর্ড উপস্থিতি সহ অ্যাপলিও ইন্টিগ্রেশন সক্ষম করুন",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "এটি ডিসকর্ডে বর্তমান অ্যাপলিও ক্রিয়াকলাপ প্রদর্শনের সম্ভাবনা সক্রিয় করবে।",
+ "Enable Applio integration with applio.org/models using flask": "ফ্লাস্ক ব্যবহার করে applio.org/models সাথে অ্যাপলিও ইন্টিগ্রেশন সক্ষম করুন",
+ "It will activate the possibility of downloading models with a click from the website.": "এটি ওয়েবসাইট থেকে একটি ক্লিকের সাথে মডেলগুলি ডাউনলোড করার সম্ভাবনা সক্রিয় করবে।",
+ "Theme": "থীম",
+ "Select the theme you want to use. (Requires restarting Applio)": "আপনি যে থিমটি ব্যবহার করতে চান তা নির্বাচন করুন। (অ্যাপলিও পুনরায় চালু করা প্রয়োজন)",
+ "Language": "ভাষা",
+ "Select the language you want to use. (Requires restarting Applio)": "আপনি যে ভাষাটি ব্যবহার করতে চান তা নির্বাচন করুন। (অ্যাপলিও পুনরায় চালু করা প্রয়োজন)",
+ "Plugin Installer": "প্লাগইন ইনস্টলার",
+ "Drag your plugin.zip to install it": "এটি ইনস্টল করতে আপনার plugin.zip টেনে আনুন",
+ "Version Checker": "সংস্করণ পরীক্ষক",
+ "Check which version of Applio is the latest to see if you need to update.": "আপনার আপডেট করার প্রয়োজন আছে কিনা তা দেখতে অ্যাপলিওর কোন সংস্করণটি সর্বশেষতম তা পরীক্ষা করে দেখুন।",
+ "Check for updates": "আপডেটের জন্য পরীক্ষা করুন"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/de_DE.json b/assets/i18n/languages/de_DE.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd0c75f0aa2b413004309a09f8270b3746ee0b8b
--- /dev/null
+++ b/assets/i18n/languages/de_DE.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Ultimatives Tool zum Klonen von Stimmen, das sorgfältig für unübertroffene Leistung, Modularität und Benutzerfreundlichkeit optimiert wurde.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Dieser Abschnitt enthält einige zusätzliche Dienstprogramme, die sich häufig in experimentellen Phasen befinden.",
+ "Output Information": "Informationen zur Ausgabe",
+ "The output information will be displayed here.": "Hier werden die Ausgabeinformationen angezeigt.",
+ "Inference": "Inferenz",
+ "Train": "Trainieren",
+ "Extra": "Extra",
+ "Merge Audios": "Audios zusammenführen",
+ "Processing": "Verarbeitung",
+ "Audio Analyzer": "Audio-Analysator",
+ "Model Information": "Modell-Informationen",
+ "Plugins": "Plugins",
+ "Download": "Herunterladen",
+ "Report a Bug": "Einen Fehler melden",
+ "Settings": "Einstellungen",
+ "Preprocess": "Vorverarbeiten",
+ "Model Name": "Modellname",
+ "Name of the new model.": "Name des neuen Modells.",
+ "Enter model name": "Modellnamen eingeben",
+ "Dataset Path": "Datensatz-Pfad",
+ "Path to the dataset folder.": "Pfad zum Datensatz-Ordner.",
+ "Refresh Datasets": "Aktualisiere den Datensatz",
+ "Dataset Creator": "Ersteller des Datensatzes",
+ "Dataset Name": "Name des Datensatzes",
+ "Name of the new dataset.": "Name des neuen Datensatzes.",
+ "Enter dataset name": "Geben Sie den Namen des Datensatzes ein",
+ "Upload Audio Dataset": "Audio-Datensatz hochladen",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Die Audiodatei wurde erfolgreich zum Datensatz hinzugefügt. Bitte klicken Sie auf die Schaltfläche \"Vorverarbeiten\".",
+ "Enter dataset path": "Datensatz-Pfad eingeben",
+ "Sampling Rate": "Samplingrate",
+ "The sampling rate of the audio files.": "Die Samplingrate der Audiodateien.",
+ "RVC Version": "RVC-Version",
+ "The RVC version of the model.": "Die RVC-Version des Modells.",
+ "Preprocess Dataset": "Datensatz vorverarbeiten",
+ "Extract": "Extrahieren",
+ "Hop Length": "Sprungweite",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Gibt die Dauer an, die das System benötigt, um zu einer signifikanten Tonhöhenänderung überzugehen. Kleinere Sprunglängen benötigen mehr Zeit für die Inferenz, führen aber tendenziell zu einer höheren Tonhöhengenauigkeit.",
+ "Batch Size": "Losgröße",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Es ist ratsam, es mit dem verfügbaren VRAM Ihrer GPU in Einklang zu bringen. Eine Einstellung von 4 bietet eine verbesserte Genauigkeit, aber eine langsamere Verarbeitung, während 8 schnellere und standardmäßige Ergebnisse liefert.",
+ "Save Every Epoch": "Speichere nach jeder Epoche",
+ "Determine at how many epochs the model will saved at.": "Legen Sie fest, bei wie vielen Epochen das Modell gespeichert wird.",
+ "Total Epoch": "Epoche insgesamt",
+ "Specifies the overall quantity of epochs for the model training process.": "Gibt die Gesamtanzahl der Epochen für den Modelltrainingsprozess an.",
+ "Pretrained": "Vortrainiert",
+ "Save Only Latest": "Nur das Neueste speichern",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Wenn Sie diese Einstellung aktivieren, werden in den G- und D-Dateien nur die neuesten Versionen gespeichert, wodurch Speicherplatz gespart wird.",
+ "Save Every Weights": "Speichern Sie alle Gewichtungen",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Mit dieser Einstellung können Sie die Gewichtungen des Modells am Ende jeder Epoche speichern.",
+ "Custom Pretrained": "Benutzerdefiniert vortrainiert",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Die Verwendung benutzerdefinierter vortrainierter Modelle kann zu besseren Ergebnissen führen, da die Auswahl der am besten geeigneten vortrainierten Modelle, die auf den jeweiligen Anwendungsfall zugeschnitten sind, die Leistung erheblich verbessern kann.",
+ "Upload Pretrained Model": "Vortrainiertes Modell hochladen",
+ "Refresh Custom Pretraineds": "Aktualisiere benutzerdefinierte vortrainierte",
+ "Pretrained Custom Settings": "Vortrainierte benutzerdefinierte Einstellungen",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Die Datei, die Sie abgelegt haben, ist keine gültige vortrainierte Datei. Bitte versuchen Sie es erneut.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Klicken Sie auf die Schaltfläche \"Aktualisieren\", um die vortrainierte Datei im Dropdown-Menü anzuzeigen.",
+ "Pretrained G Path": "Benutzerdefinierter vortrainierter G Pfad",
+ "Pretrained D Path": "Benutzerdefinierter vortrainierter D Pfad",
+ "GPU Settings": "GPU-Einstellungen",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Legt erweiterte GPU-Einstellungen fest, die für Benutzer mit besserer GPU-Architektur empfohlen werden.",
+ "GPU Custom Settings": "Benutzerdefinierte GPU-Einstellungen",
+ "GPU Number": "GPU-Nummer",
+ "0 to ∞ separated by -": "0 bis ∞ getrennt durch -",
+ "GPU Information": "GPU-Informationen",
+ "Pitch Guidance": "Tonhöhen-Führung",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Durch den Einsatz von Tonhöhenführung wird es möglich, die Intonation der Originalstimme, einschließlich ihrer Tonhöhe, zu spiegeln. Diese Funktion ist besonders wertvoll für das Singen und andere Szenarien, in denen die Beibehaltung der ursprünglichen Melodie oder des Tonhöhenmusters unerlässlich ist.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Verwenden Sie vortrainierte Modelle, wenn Sie Ihre eigenen trainieren. Dieser Ansatz verkürzt die Schulungsdauer und verbessert die Gesamtqualität.",
+ "Extract Features": "Extrahieren von Merkmalen",
+ "Start Training": "Training starten",
+ "Generate Index": "Index generieren",
+ "Voice Model": "Sprach-Modell",
+ "Select the voice model to use for the conversion.": "Wählen Sie das Sprachmodell aus, das für die Konvertierung verwendet werden soll.",
+ "Index File": "Index-Datei",
+ "Select the index file to use for the conversion.": "Wählen Sie die Indexdatei aus, die für die Konvertierung verwendet werden soll.",
+ "Refresh": "Aktualisieren",
+ "Unload Voice": "Stimme entladen",
+ "Single": "Einzeln",
+ "Upload Audio": "Audio hochladen",
+ "Select Audio": "Wählen Sie ein Audio",
+ "Select the audio to convert.": "Wählen Sie das zu konvertierende Audio aus.",
+ "Advanced Settings": "Erweiterte Einstellungen",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Ausgaben löschen (Löscht alle Audios in assets/audios)",
+ "Custom Output Path": "Benutzerdefinierter Ausgabepfad",
+ "Output Path": "Ausgabepfad",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Der Pfad, in dem das Ausgabeaudio gespeichert wird, standardmäßig in assets/audios/output.wav",
+ "Split Audio": "Audio aufteilen",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Teilen Sie die Audiodaten für Rückschlüsse in Blöcke auf, um in einigen Fällen bessere Ergebnisse zu erzielen.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Wenden Sie ein sanftes Autotune auf Ihre Inferenzen an, das für Gesangskonvertierungen empfohlen wird.",
+ "Clean Audio": "Audio bereinigen",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Bereinigen Sie Ihre Audioausgabe mithilfe von Rauscherkennungsalgorithmen, die für gesprochene Audios empfohlen werden.",
+ "Clean Strength": "Bereinigungsstärke",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Stellen Sie den Bereinigungsstärke auf den gewünschten Wert ein, je mehr Sie ihn erhöhen, desto mehr wird bereinigt, aber es ist möglich, dass der Ton stärker komprimiert wird.",
+ "Pitch": "Tonhöhe",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Stellen Sie die Tonhöhe des Audios ein, je höher der Wert, desto höher die Tonhöhe.",
+ "Filter Radius": "Filter-Radius",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Wenn die Zahl größer oder gleich drei ist, kann die Verwendung einer Medianfilterung für die gesammelten Tonergebnisse die Atmung verringern.",
+ "Search Feature Ratio": "Such-Merkmal-Verhältnis",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Einfluss, der von der Indexdatei ausgeübt wird; Ein höherer Wert entspricht einem größeren Einfluss. Wenn Sie sich jedoch für niedrigere Werte entscheiden, können Sie Artefakte im Audiomaterial abschwächen.",
+ "Volume Envelope": "Lautstärke-Hüllkurve",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Ersetzen oder überblenden Sie die Lautstärke-Hüllkurve des Ausgangs. Je näher das Verhältnis an 1 liegt, desto mehr wird die Ausgangshüllkurve verwendet.",
+ "Protect Voiceless Consonants": "Schützen Sie stimmlose Konsonanten",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Schützen Sie eindeutige Konsonanten und Atemgeräusche, um elektroakustisches Reißen und andere Artefakte zu vermeiden. Das Ziehen des Parameters auf den Maximalwert von 0,5 bietet einen umfassenden Schutz. Das Verringern dieses Werts kann jedoch den Umfang des Schutzes verringern und gleichzeitig den Indizierungseffekt möglicherweise abschwächen.",
+ "Pitch extraction algorithm": "Algorithmus zur Tonhöhenextraktion",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Tonhöhenextraktionsalgorithmus, der für die Audiokonvertierung verwendet werden soll. Der Standardalgorithmus ist rmvpe, der in den meisten Fällen empfohlen wird.",
+ "Convert": "Umwandeln",
+ "Export Audio": "Audio exportieren",
+ "Batch": "Charge",
+ "Input Folder": "Eingabe-Ordner",
+ "Select the folder containing the audios to convert.": "Wählen Sie den Ordner aus, der die zu konvertierenden Audios enthält.",
+ "Enter input path": "Eingabepfad eingeben",
+ "Output Folder": "Ausgabe-Ordner",
+ "Select the folder where the output audios will be saved.": "Wählen Sie den Ordner aus, in dem die ausgegebenen Audios gespeichert werden sollen.",
+ "Enter output path": "Ausgabepfad eingeben",
+ "Get information about the audio": "Abrufen von Informationen zum Audio",
+ "Information about the audio file": "Informationen zur Audiodatei",
+ "Waiting for information...": "Warten auf Informationen...",
+ "## Voice Blender": "## Voice Blender",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Wählen Sie zwei Stimmmodelle aus, legen Sie den gewünschten Überblendungsprozentsatz fest und mischen Sie sie zu einer völlig neuen Stimme.",
+ "Voice Blender": "Voice Blender",
+ "Drag and drop your model here": "Ziehen Sie Ihr Modell per Drag & Drop hierher",
+ "You can also use a custom path.": "Sie können auch einen benutzerdefinierten Pfad verwenden.",
+ "Blend Ratio": "Mischungsverhältnis",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Wenn Sie die Position mehr auf die eine oder andere Seite anpassen, wird das Modell dem ersten oder zweiten ähnlicher.",
+ "Fusion": "Verschmelzen",
+ "Path to Model": "Pfad zum Modell",
+ "Enter path to model": "Pfad zum Modell eingeben",
+ "Model information to be placed": "Zu platzierende Modellinformationen",
+ "Inroduce the model information": "Einfügen der Modellinformationen",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Die Informationen, die in das Modell eingefügt werden sollen (Sie können das Feld leer lassen oder etwas anderes einfügen).",
+ "View model information": "Anzeigen von Modellinformationen",
+ "Introduce the model pth path": "Einfügen des pth Pfad des Modells",
+ "View": "Ansehen",
+ "Model extraction": "Modell-Extraktion",
+ "Model conversion": "Modell-Konvertierung",
+ "Pth file": "Pth-Datei",
+ "Output of the pth file": "Ausgabe der pth-Datei",
+ "# How to Report an Issue on GitHub": "# So melden Sie ein Problem auf GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Klicken Sie unten auf die Schaltfläche \"Bildschirm aufzeichnen\", um mit der Aufzeichnung des aufgetretenen Problems zu beginnen.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Wenn Sie mit der Aufnahme des Problems fertig sind, klicken Sie auf die Schaltfläche \"Aufnahme beenden\" (dieselbe Schaltfläche, aber die Beschriftung ändert sich, je nachdem, ob Sie aktiv aufnehmen oder nicht).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Gehen Sie zu [GitHub Issues](https://github.com/IAHispano/Applio/issues) und klicken Sie auf die Schaltfläche \"New Issue\".",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Füllen Sie die bereitgestellte Problemvorlage aus, stellen Sie sicher, dass Sie die erforderlichen Details angeben, und verwenden Sie den Abschnitt \"Assets\", um die aufgezeichnete Datei aus dem vorherigen Schritt hochzuladen.",
+ "Record Screen": "Bildschirm aufzeichnen",
+ "Record": "Aufzeichnen",
+ "Stop Recording": "Aufzeichnung beenden",
+ "Introduce the model .pth path": "Einfügen des .pth Pfad des Modells",
+ "See Model Information": "Siehe Modellinformationen",
+ "## Download Model": "## Modell herunterladen",
+ "Model Link": "Modell-Link",
+ "Introduce the model link": "Einfügen des Modell-Links",
+ "Download Model": "Modell herunterladen",
+ "## Drop files": "## Dateien ablegen",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Ziehen Sie Ihre .pth und .index Datei in diesen Bereich. Ziehen Sie das eine und dann das andere.",
+ "TTS Voices": "TTS-Stimmen",
+ "Select the TTS voice to use for the conversion.": "Wählen Sie die TTS-Stimme aus, die für die Konvertierung verwendet werden soll.",
+ "Text to Synthesize": "Zu synthetisierender Text",
+ "Enter the text to synthesize.": "Geben Sie den zu synthetisierenden Text ein.",
+ "Or you can upload a .txt file": "Oder Sie können eine .txt Datei hochladen",
+ "Enter text to synthesize": "Geben Sie den zu synthetisierenden Text ein",
+ "Output Path for TTS Audio": "Ausgabepfad für das TTS-Audio",
+ "Output Path for RVC Audio": "Ausgabepfad für das RVC-Audio",
+ "Enable Applio integration with Discord presence": "Aktivieren Sie die Applio-Integration mit Discord-Presence",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Es aktiviert die Möglichkeit, die aktuelle Applio-Aktivität in Discord anzuzeigen.",
+ "Enable Applio integration with applio.org/models using flask": "Aktivieren der Applio-Integration mit applio.org/models mithilfe von flask",
+ "It will activate the possibility of downloading models with a click from the website.": "Es aktiviert die Möglichkeit, Modelle mit einem Klick von der Website herunterzuladen.",
+ "Theme": "Design",
+ "Select the theme you want to use. (Requires restarting Applio)": "Wählen Sie das Design aus, das Sie verwenden möchten. (Erfordert einen Neustart von Applio)",
+ "Language": "Sprache",
+ "Select the language you want to use. (Requires restarting Applio)": "Wählen Sie die Sprache aus, die Sie verwenden möchten. (Erfordert einen Neustart von Applio)",
+ "Plugin Installer": "Plugin-Installer",
+ "Drag your plugin.zip to install it": "Ziehen Sie Ihre plugin.zip, um sie zu installieren",
+ "Version Checker": "Versions-Checker",
+ "Check which version of Applio is the latest to see if you need to update.": "Überprüfen Sie, welche Version von Applio die neueste ist, um zu sehen, ob Sie ein Update benötigen.",
+ "Check for updates": "Nach Updates suchen"
+}
diff --git a/assets/i18n/languages/en_US.json b/assets/i18n/languages/en_US.json
new file mode 100644
index 0000000000000000000000000000000000000000..23189320a5cfc9dffe439412886d51c6fa353351
--- /dev/null
+++ b/assets/i18n/languages/en_US.json
@@ -0,0 +1,224 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.",
+ "This section contains some extra utilities that often may be in experimental phases.": "This section contains some extra utilities that often may be in experimental phases.",
+ "Output Information": "Output Information",
+ "The output information will be displayed here.": "The output information will be displayed here.",
+
+ "Inference": "Inference",
+ "Train": "Train",
+ "Extra": "Extra",
+ "Merge Audios": "Merge Audios",
+ "Processing": "Processing",
+ "Audio Analyzer": "Audio Analyzer",
+ "Model Information": "Model Information",
+ "Plugins": "Plugins",
+ "Download": "Download",
+ "Report a Bug": "Report a Bug",
+ "Settings": "Settings",
+
+ "Preprocess": "Preprocess",
+ "Model Name": "Model Name",
+ "Name of the new model.": "Name of the new model.",
+ "Enter model name": "Enter model name",
+ "Dataset Path": "Dataset Path",
+ "Path to the dataset folder.": "Path to the dataset folder.",
+ "Refresh Datasets": "Refresh Datasets",
+ "Dataset Creator": "Dataset Creator",
+ "Dataset Name": "Dataset Name",
+ "Name of the new dataset.": "Name of the new dataset.",
+ "Enter dataset name": "Enter dataset name",
+ "Upload Audio Dataset": "Upload Audio Dataset",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "The audio file has been successfully added to the dataset. Please click the preprocess button.",
+ "Enter dataset path": "Enter dataset path",
+ "Sampling Rate": "Sampling Rate",
+ "The sampling rate of the audio files.": "The sampling rate of the audio files.",
+ "RVC Version": "RVC Version",
+ "The RVC version of the model.": "The RVC version of the model.",
+ "Preprocess Dataset": "Preprocess Dataset",
+
+ "Embedder Model": "Embedder Model",
+ "Model used for learning speaker embedding.": "Model used for learning speaker embedding.",
+ "Extract": "Extract",
+ "Hop Length": "Hop Length",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.",
+ "Batch Size": "Batch Size",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.",
+ "Save Every Epoch": "Save Every Epoch",
+ "Determine at how many epochs the model will saved at.": "Determine at how many epochs the model will saved at.",
+ "Total Epoch": "Total Epoch",
+ "Specifies the overall quantity of epochs for the model training process.": "Specifies the overall quantity of epochs for the model training process.",
+ "Pretrained": "Pretrained",
+ "Save Only Latest": "Save Only Latest",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.",
+ "Save Every Weights": "Save Every Weights",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "This setting enables you to save the weights of the model at the conclusion of each epoch.",
+ "Custom Pretrained": "Custom Pretrained",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.",
+ "Upload Pretrained Model": "Upload Pretrained Model",
+ "Refresh Custom Pretraineds": "Refresh Custom Pretraineds",
+ "Pretrained Custom Settings": "Pretrained Custom Settings",
+ "The file you dropped is not a valid pretrained file. Please try again.": "The file you dropped is not a valid pretrained file. Please try again.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Click the refresh button to see the pretrained file in the dropdown menu.",
+ "Pretrained G Path": "Custom Pretrained G",
+ "Pretrained D Path": "Custom Pretrained D",
+ "GPU Settings": "GPU Settings",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Sets advanced GPU settings, recommended for users with better GPU architecture.",
+ "GPU Custom Settings": "GPU Custom Settings",
+ "GPU Number": "GPU Number",
+ "0 to ∞ separated by -": "0 to ∞ separated by -",
+ "GPU Information": "GPU Information",
+ "Pitch Guidance": "Pitch Guidance",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.",
+ "Extract Features": "Extract Features",
+ "Overtraining Detector": "Overtraining Detector",
+ "Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data.": "Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data.",
+ "Overtraining Detector Settings": "Overtraining Detector Settings",
+ "Overtraining Threshold": "Overtraining Threshold",
+ "Set the maximum number of epochs you want your model to stop training if no improvement is detected.": "Set the maximum number of epochs you want your model to stop training if no improvement is detected.",
+ "Sync Graph": "Sync Graph",
+ "Synchronize the graph of the tensorbaord. Only enable this setting if you are training a new model.": "Synchronize the graph of the tensorbaord. Only enable this setting if you are training a new model.",
+
+ "Start Training": "Start Training",
+ "Stop Training & Restart Applio": "Stop Training & Restart Applio",
+ "Generate Index": "Generate Index",
+
+ "Export Model": "Export Model",
+ "The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive.": "The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive.",
+ "Exported Pth file": "Exported Pth file",
+ "Exported Index file": "Exported Index file",
+ "Select the pth file to be exported": "Select the pth file to be exported",
+ "Select the index file to be exported": "Select the index file to be exported",
+ "Upload": "Upload",
+
+ "Voice Model": "Voice Model",
+ "Select the voice model to use for the conversion.": "Select the voice model to use for the conversion.",
+ "Index File": "Index File",
+ "Select the index file to use for the conversion.": "Select the index file to use for the conversion.",
+ "Refresh": "Refresh",
+ "Unload Voice": "Unload Voice",
+
+ "Single": "Single",
+ "Upload Audio": "Upload Audio",
+ "Select Audio": "Select Audio",
+ "Select the audio to convert.": "Select the audio to convert.",
+ "Advanced Settings": "Advanced Settings",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Clear Outputs (Deletes all audios in assets/audios)",
+ "Custom Output Path": "Custom Output Path",
+ "Output Path": "Output Path",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "The path where the output audio will be saved, by default in assets/audios/output.wav",
+ "Split Audio": "Split Audio",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Split the audio into chunks for inference to obtain better results in some cases.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Apply a soft autotune to your inferences, recommended for singing conversions.",
+ "Clean Audio": "Clean Audio",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Clean your audio output using noise detection algorithms, recommended for speaking audios.",
+ "Clean Strength": "Clean Strength",
+ "Upscale Audio": "Upscale Audio",
+ "Upscale the audio to a higher quality, recommended for low-quality audios. (It could take longer to process the audio)": "Upscale the audio to a higher quality, recommended for low-quality audios. (It could take longer to process the audio)",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.",
+ "Pitch": "Pitch",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Set the pitch of the audio, the higher the value, the higher the pitch.",
+ "Filter Radius": "Filter Radius",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.",
+ "Search Feature Ratio": "Search Feature Ratio",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.",
+ "Volume Envelope": "Volume Envelope",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.",
+ "Protect Voiceless Consonants": "Protect Voiceless Consonants",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.",
+ "Pitch extraction algorithm": "Pitch extraction algorithm",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.",
+
+ "Convert": "Convert",
+ "Export Audio": "Export Audio",
+
+ "Batch": "Batch",
+ "Input Folder": "Input Folder",
+ "Select the folder containing the audios to convert.": "Select the folder containing the audios to convert.",
+ "Enter input path": "Enter input path",
+ "Output Folder": "Output Folder",
+ "Select the folder where the output audios will be saved.": "Select the folder where the output audios will be saved.",
+ "Enter output path": "Enter output path",
+
+ "Get information about the audio": "Get information about the audio",
+
+ "## Voice Blender": "## Voice Blender",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.",
+ "Voice Blender": "Voice Blender",
+ "Drag and drop your model here": "Drag and drop your model here",
+ "You can also use a custom path.": "You can also use a custom path.",
+ "Blend Ratio": "Blend Ratio",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Adjusting the position more towards one side or the other will make the model more similar to the first or second.",
+ "Fusion": "Fusion",
+
+ "Path to Model": "Path to Model",
+ "Enter path to model": "Enter path to model",
+ "Model information to be placed": "Model information to be placed",
+ "Inroduce the model information": "Inroduce the model information",
+ "The information to be placed in the model (You can leave it blank or put anything).": "The information to be placed in the model (You can leave it blank or put anything).",
+ "View model information": "View model information",
+ "Introduce the model pth path": "Introduce the model pth path",
+ "View": "View",
+ "Model extraction": "Model extraction",
+ "Model conversion": "Model conversion",
+ "Pth file": "Pth file",
+ "Output of the pth file": "Output of the pth file",
+
+ "# How to Report an Issue on GitHub": "# How to Report an Issue on GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.",
+
+ "Record Screen": "Record Screen",
+ "Record": "Record",
+ "Stop Recording": "Stop Recording",
+
+ "Introduce the model .pth path": "Introduce the model .pth path",
+ "See Model Information": "See Model Information",
+
+ "## Download Model": "## Download Model",
+ "Model Link": "Model Link",
+ "Introduce the model link": "Introduce the model link",
+ "Download Model": "Download Model",
+ "## Drop files": "## Drop files",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Drag your .pth file and .index file into this space. Drag one and then the other.",
+ "## Search Model": "## Search Model",
+ "Search": "Search",
+ "Introduce the model name to search.": "Introduce the model name to search.",
+ "We couldn't find models by that name.": "We couldn't find models by that name.",
+ "## Download Pretrained Models": "## Download Pretrained Models",
+ "Select the pretrained model you want to download.": "Select the pretrained model you want to download.",
+ "And select the sampling rate": "And select the sampling rate.",
+
+ "TTS Voices": "TTS Voices",
+ "TTS Speed": "TTS Speed",
+ "Increase or decrease TTS speed.": "Increase or decrease TTS speed.",
+ "Select the TTS voice to use for the conversion.": "Select the TTS voice to use for the conversion.",
+ "Text to Synthesize": "Text to Synthesize",
+ "Enter the text to synthesize.": "Enter the text to synthesize.",
+ "Or you can upload a .txt file": "Or you can upload a .txt file",
+ "Enter text to synthesize": "Enter text to synthesize",
+ "Output Path for TTS Audio": "Output Path for TTS Audio",
+ "Output Path for RVC Audio": "Output Path for RVC Audio",
+
+ "Enable Applio integration with Discord presence": "Enable Applio integration with Discord presence",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "It will activate the possibility of displaying the current Applio activity in Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Enable Applio integration with applio.org/models using flask",
+ "It will activate the possibility of downloading models with a click from the website.": "It will activate the possibility of downloading models with a click from the website.",
+ "Enable fake GPU": "Enable fake GPU",
+ "Training is currently unsupported due to the absence of a GPU. To activate the training tab, navigate to the settings tab and enable the 'Fake GPU' option.": "Training is currently unsupported due to the absence of a GPU. To activate the training tab, navigate to the settings tab and enable the 'Fake GPU' option.",
+ "Activates the train tab. However, please note that this device lacks GPU capabilities, hence training is not supported. This option is only for testing purposes. (This option will restart Applio)": "Activates the train tab. However, please note that this device lacks GPU capabilities, hence training is not supported. This option is only for testing purposes. (This option will restart Applio)",
+ "Theme": "Theme",
+ "Select the theme you want to use. (Requires restarting Applio)": "Select the theme you want to use. (Requires restarting Applio)",
+ "Language": "Language",
+ "Select the language you want to use. (Requires restarting Applio)": "Select the language you want to use. (Requires restarting Applio)",
+
+ "Plugin Installer": "Plugin Installer",
+ "Drag your plugin.zip to install it": "Drag your plugin.zip to install it",
+
+ "Version Checker": "Version Checker",
+ "Check which version of Applio is the latest to see if you need to update.": "Check which version of Applio is the latest to see if you need to update.",
+ "Check for updates": "Check for updates"
+}
diff --git a/assets/i18n/languages/es_ES.json b/assets/i18n/languages/es_ES.json
new file mode 100644
index 0000000000000000000000000000000000000000..700d0c28e92cf4bf5b1f162b060a8f6fb6f34795
--- /dev/null
+++ b/assets/i18n/languages/es_ES.json
@@ -0,0 +1,205 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "La herramienta de clonación de voz definitiva, meticulosamente optimizada para una potencia, modularidad y experiencia de uso inigualables.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Esta sección contiene algunas utilidades adicionales que a menudo pueden estar en fases experimentales.",
+ "Output Information": "Información de salida",
+ "The output information will be displayed here.": "La información de salida se mostrará aquí.",
+ "Inference": "Inferencia",
+ "Train": "Entrenar",
+ "Extra": "Extra",
+ "Merge Audios": "Fusionar audios",
+ "Processing": "Procesamiento",
+ "Audio Analyzer": "Analizador de audio",
+ "Model Information": "Información del modelo",
+ "Plugins": "Complementos",
+ "Download": "Descargar",
+ "Report a Bug": "Informar de un error",
+ "Settings": "Configuración",
+ "Preprocess": "Preprocesar",
+ "Model Name": "Nombre del modelo",
+ "Name of the new model.": "Nombre del nuevo modelo.",
+ "Enter model name": "Introduzca el nombre del modelo",
+ "Dataset Path": "Ruta del dataset",
+ "Path to the dataset folder.": "Ruta de acceso al dataset.",
+ "Refresh Datasets": "Actualizar datasets",
+ "Dataset Creator": "Creador de datasets",
+ "Dataset Name": "Nombre del dataset",
+ "Name of the new dataset.": "Nombre del nuevo dataset.",
+ "Enter dataset name": "Introduzca el nombre del dataset",
+ "Upload Audio Dataset": "Cargar audio del dataset",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "El archivo de audio se ha agregado correctamente dataset. Haga clic en el botón de preprocesamiento.",
+ "Enter dataset path": "Introduzca la ruta del dataset",
+ "Sampling Rate": "Frecuencia de muestreo",
+ "The sampling rate of the audio files.": "La frecuencia de muestreo de los archivos de audio.",
+ "RVC Version": "Versión RVC",
+ "The RVC version of the model.": "La versión RVC del modelo.",
+ "Preprocess Dataset": "Conjunto de datos de preprocesamiento",
+ "Embedder Model": "Modelo de incrustación",
+ "Model used for learning speaker embedding.": "Modelo utilizado para aprender la incrustación del hablante.",
+ "Hop Length": "Longitud del salto",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Denota el tiempo que tarda el sistema en realizar la transición a un cambio de tono significativo. Las longitudes de salto más pequeñas requieren más tiempo para la inferencia, pero tienden a producir una mayor precisión de tono.",
+ "Batch Size": "Tamaño del lote",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Es recomendable alinearlo con la VRAM disponible de tu GPU. Un ajuste de 4 ofrece una precisión mejorada pero un procesamiento más lento, mientras que 8 proporciona resultados más rápidos y estándar.",
+ "Save Every Epoch": "Frecuencia de guardado",
+ "Determine at how many epochs the model will saved at.": "Determine en cuántas épocas se guardará el modelo.",
+ "Total Epoch": "Épocas",
+ "Specifies the overall quantity of epochs for the model training process.": "Especifica la cantidad total de épocas para el proceso de entrenamiento del modelo.",
+ "Pretrained": "Preentrenado",
+ "Save Only Latest": "Guarde solo lo último",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Al habilitar esta configuración, los archivos G y D guardarán solo sus versiones más recientes, lo que ahorrará espacio de almacenamiento de manera efectiva.",
+ "Save Every Weights": "Ahorre todos los pesos",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Este ajuste le permite guardar los pesos del modelo al final de cada época.",
+ "Custom Pretrained": "Preentrenado personalizado",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "El uso de modelos preentrenados personalizados puede conducir a resultados superiores, ya que la selección de los modelos preentrenados más adecuados adaptados al caso de uso específico puede mejorar significativamente el rendimiento.",
+ "Upload Pretrained Model": "Cargar modelo previamente entrenado",
+ "Refresh Custom Pretraineds": "Actualizar preentrenados personalizados",
+ "Pretrained Custom Settings": "Configuración personalizada previamente entrenada",
+ "The file you dropped is not a valid pretrained file. Please try again.": "El archivo que has subido no es un archivo preentrenado válido. Por favor, inténtelo de nuevo.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Haga clic en el botón de actualización para ver el archivo previamente entrenado en el menú desplegable.",
+ "Pretrained G Path": "G preentrenado personalizado",
+ "Pretrained D Path": "D preentrenado personalizado",
+ "GPU Settings": "Configuración de GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Establece la configuración avanzada del GPU, recomendada para usuarios con una mejor arquitectura de GPU.",
+ "GPU Custom Settings": "Configuración personalizada de la GPU",
+ "GPU Number": "Número de GPU",
+ "0 to ∞ separated by -": "0 a ∞ separados por -",
+ "GPU Information": "Información de GPU",
+ "Pitch Guidance": "Guía de tono",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Al emplear la guía de tono, es factible reflejar la entonación de la voz original, incluido su tono. Esta característica es particularmente valiosa para cantar y otros escenarios en los que es esencial preservar la melodía original o el patrón de tono.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Utilice modelos preentrenados al entrenar los suyos propios. Este enfoque reduce la duración del entrenamiento y mejora la calidad general.",
+ "Extract Features": "Extraer características",
+ "Overtraining Detector": "Detector de Sobreentrenamiento",
+ "Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data.": "Detectar el sobreentrenamiento para evitar que el modelo aprenda demasiado bien los datos de entrenamiento y pierda la capacidad de generalizar a nuevos datos.",
+ "Overtraining Detector Settings": "Configuraciones del Detector de Sobreentrenamiento",
+ "Overtraining Threshold": "Umbral de Sobreentrenamiento",
+ "Set the maximum number of epochs you want your model to stop training if no improvement is detected.": "Establezca el número máximo de épocas para que su modelo deje de entrenar si no se detecta ninguna mejora.",
+ "Sync Graph": "Sincronizar gráfico",
+ "Synchronize the graph of the tensorbaord. Only enable this setting if you are training a new model.": "Sincronice el gráfico del tensorbaord. Solo active esta configuración si está entrenando un nuevo modelo.",
+ "Start Training": "Empezar a entrenar",
+ "Stop Training & Restart Applio": "Detener el entrenamiento y reiniciar Applio",
+ "Generate Index": "Generar índice",
+ "Export Model": "Exportar modelo",
+ "The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive.": "El botón 'Subir' es solo para Google Colab: Sube los archivos exportados a la carpeta ApplioExported en tu Google Drive.",
+ "Exported Pth file": "Archivo Pth exportado",
+ "Exported Index file": "Archivo de índice exportado",
+ "Select the pth file to be exported": "Seleccione el archivo pth que se va a exportar",
+ "Select the index file to be exported": "Seleccione el archivo de índice que se va a exportar",
+ "Upload": "Subir",
+ "Voice Model": "Modelo de voz",
+ "Select the voice model to use for the conversion.": "Seleccione el modelo de voz que desea utilizar para la conversión.",
+ "Index File": "Archivo de índice",
+ "Select the index file to use for the conversion.": "Seleccione el archivo de índice que desea utilizar para la conversión.",
+ "Refresh": "Actualizar",
+ "Unload Voice": "Descargar voz",
+ "Single": "Solo",
+ "Upload Audio": "Subir audio",
+ "Select Audio": "Seleccionar audio",
+ "Select the audio to convert.": "Seleccione el audio que desea convertir.",
+ "Advanced Settings": "Ajustes avanzados",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Borrar salidas (elimina todos los audios de los assets/audios)",
+ "Custom Output Path": "Ruta de salida personalizada",
+ "Output Path": "Ruta de salida",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "La ruta donde se guardará el audio de salida, de forma predeterminada en assets/audios/output.wav",
+ "Split Audio": "Audio dividido",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Divida el audio en fragmentos para inferir y obtener mejores resultados en algunos casos.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Aplique un autotune suave a sus inferencias, recomendado para conversiones de canto.",
+ "Clean Audio": "Audio limpio",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Limpie su salida de audio utilizando algoritmos de detección de ruido, recomendados para audios hablados.",
+ "Clean Strength": "Fuerza de limpieza",
+ "Upscale Audio": "Audio de alta calidad",
+ "Upscale the audio to a higher quality, recommended for low-quality audios. (It could take longer to process the audio)": "Mejore la calidad del audio a una calidad superior, recomendado para audios de baja calidad. (Podría tardar más en procesar el audio)",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Establezca el nivel de limpieza en el audio que desee, cuanto más lo aumente, más se limpiará, pero es posible que el audio esté más comprimido.",
+ "Pitch": "Tono",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Establezca el tono del audio, cuanto mayor sea el valor, mayor será el tono.",
+ "Filter Radius": "Radio del filtro",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Si el número es mayor o igual a tres, el empleo de filtrado de mediana en los resultados de tono recopilados tiene el potencial de disminuir la respiración.",
+ "Search Feature Ratio": "Relación de características de búsqueda",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Influencia ejercida por el fichero índice; Un valor más alto corresponde a una mayor influencia. Sin embargo, optar por valores más bajos puede ayudar a mitigar los artefactos presentes en el audio.",
+ "Volume Envelope": "Envoltura de volumen",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Sustituya o mezcle con la envoltura de volumen de la salida. Cuanto más se acerque la relación a 1, más se empleará la envoltura de salida.",
+ "Protect Voiceless Consonants": "Proteger las consonantes sordas",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Proteja las distintas consonantes y sonidos respiratorios para evitar desgarros electroacústicos y otros artefactos. Llevar el parámetro a su valor máximo de 0,5 ofrece una protección completa. Sin embargo, la reducción de este valor podría disminuir el alcance de la protección y, al mismo tiempo, mitigar el efecto de indexación.",
+ "Pitch extraction algorithm": "Algoritmo de extracción de tono",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Algoritmo de extracción de tono que se utilizará para la conversión de audio. El algoritmo predeterminado es rmvpe, que se recomienda para la mayoría de los casos.",
+ "Convert": "Convertir",
+ "Export Audio": "Exportar audio",
+ "Batch": "Lote",
+ "Input Folder": "Carpeta de entrada",
+ "Select the folder containing the audios to convert.": "Seleccione la carpeta que contiene los audios que desea convertir.",
+ "Enter input path": "Introduzca la ruta de entrada",
+ "Output Folder": "Carpeta de salida",
+ "Select the folder where the output audios will be saved.": "Seleccione la carpeta donde se guardarán los audios de salida.",
+ "Enter output path": "Introduzca la ruta de salida",
+ "Get information about the audio": "Obtener información sobre el audio",
+ "Information about the audio file": "Información sobre el archivo de audio",
+ "Waiting for information...": "A la espera de información...",
+ "## Voice Blender": "## Mezclador de voz",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Seleccione dos modelos de voz, establezca el porcentaje de mezcla deseado y combínelos en una voz completamente nueva.",
+ "Voice Blender": "Mezclador de voz",
+ "Drag and drop your model here": "Arrastra y suelta tu modelo aquí",
+ "You can also use a custom path.": "También puede utilizar una ruta de acceso personalizada.",
+ "Blend Ratio": "Relación de mezcla",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Ajustar la posición más hacia un lado u otro hará que el modelo se parezca más al primero o al segundo.",
+ "Fusion": "Fusión",
+ "Path to Model": "Ruta de acceso al modelo",
+ "Enter path to model": "Introduzca la ruta al modelo",
+ "Model information to be placed": "Información del modelo que se va a colocar",
+ "Inroduce the model information": "Introduzca la información del modelo",
+ "The information to be placed in the model (You can leave it blank or put anything).": "La información que se va a colocar en el modelo (Puedes dejarlo en blanco o poner cualquier cosa).",
+ "View model information": "Ver información del modelo",
+ "Introduce the model pth path": "Introducción de la ruta pth del modelo",
+ "View": "Vista",
+ "Model extraction": "Extracción de modelos",
+ "Model conversion": "Conversión de modelos",
+ "Pth file": "Archivo Pth",
+ "Output of the pth file": "Salida del fichero pth",
+ "# How to Report an Issue on GitHub": "# Cómo reportar un problema en GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Haga clic en el botón 'Grabar pantalla' a continuación para comenzar a grabar el problema que está experimentando.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Una vez que haya terminado de grabar el problema, haga clic en el botón 'Detener grabación' (el mismo botón, pero la etiqueta cambia dependiendo de si está grabando activamente o no).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Vaya a [Problemas de GitHub](https://github.com/IAHispano/Applio/issues) y haga clic en el botón 'Nuevo problema'.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Complete la plantilla de problema proporcionada, asegurándose de incluir los detalles según sea necesario, y utilice la sección de activos para cargar el archivo grabado del paso anterior.",
+ "Record Screen": "Grabar pantalla",
+ "Record": "Grabar",
+ "Stop Recording": "Detener la grabación",
+ "Introduce the model .pth path": "Introducción de la ruta de acceso .pth del modelo",
+ "See Model Information": "Ver información del modelo",
+ "## Download Model": "## Descargar modelo",
+ "Model Link": "Enlace de modelo",
+ "Introduce the model link": "Introducir el enlace del modelo",
+ "Download Model": "Descargar modelo",
+ "## Drop files": "## Soltar archivos",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Arrastre el archivo .pth y el archivo .index a este espacio. Arrastre uno y luego el otro.",
+ "## Search Model": "## Buscar modelo",
+ "Search": "Buscar",
+ "Introduce the model name to search.": "Introduzca el nombre del modelo a buscar.",
+ "We couldn't find models by that name.": "No pudimos encontrar modelos con ese nombre.",
+ "## Download Pretrained Models": "## Descargar modelos preentrenados",
+ "Select the pretrained model you want to download.": "Seleccione el modelo preentrenado que desea descargar.",
+ "And select the sampling rate.": "Y seleccione la frecuencia de muestreo.",
+ "TTS Voices": "Voces TTS",
+ "TTS Speed": "Velocidad TTS",
+ "Increase or decrease TTS speed": "Aumentar o disminuir la velocidad de TTS",
+ "Select the TTS voice to use for the conversion.": "Seleccione la voz TTS que se utilizará para la conversión.",
+ "Text to Synthesize": "Texto para sintetizar",
+ "Enter the text to synthesize.": "Introduzca el texto que desea sintetizar.",
+ "Or you can upload a .txt file": "O bien, puede cargar un archivo .txt",
+ "Enter text to synthesize": "Introduzca el texto que desea sintetizar",
+ "Output Path for TTS Audio": "Ruta de salida para audio TTS",
+ "Output Path for RVC Audio": "Ruta de salida para audio RVC",
+ "Enable Applio integration with Discord presence": "Habilitar la integración de Applio con la presencia de Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Activará la posibilidad de mostrar la actividad actual de Applio en Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Habilite la integración de Applio con applio.org/models usando flask",
+ "It will activate the possibility of downloading models with a click from the website.": "Activará la posibilidad de descargar modelos con un clic desde el sitio web.",
+ "Enable fake GPU": "Habilitar GPU falsa",
+ "Training is currently unsupported due to the absence of a GPU. To activate the training tab, navigate to the settings tab and enable the 'Fake GPU' option.": "Actualmente no se admite el entrenamiento debido a la ausencia de una GPU. Para activar la pestaña de entrenamiento, vaya a la pestaña de configuración y habilite la opción 'GPU falsa'.",
+ "Activates the train tab. However, please note that this device lacks GPU capabilities, hence training is not supported. This option is only for testing purposes. (This option will restart Applio)": "Activa la pestaña de entrenamiento. Sin embargo, tenga en cuenta que este dispositivo carece de capacidades de GPU, por lo que no se admite el entrenamiento. Esta opción es solo para fines de prueba. (Esta opción reiniciará Applio)",
+ "Theme": "Tema",
+ "Select the theme you want to use. (Requires restarting Applio)": "Seleccione el tema que desea utilizar. (Requiere reiniciar Applio)",
+ "Language": "Idioma",
+ "Select the language you want to use. (Requires restarting Applio)": "Seleccione el idioma que desea utilizar. (Requiere reiniciar Applio)",
+ "Plugin Installer": "Instalador de complementos",
+ "Drag your plugin.zip to install it": "Arrastra tu plugin.zip para instalarlo",
+ "Version Checker": "Comprobador de versiones",
+ "Check which version of Applio is the latest to see if you need to update.": "Compruebe qué versión de Applio es la más reciente para ver si necesita actualizar.",
+ "Check for updates": "Buscar actualizaciones"
+}
diff --git a/assets/i18n/languages/fa_FA.json b/assets/i18n/languages/fa_FA.json
new file mode 100644
index 0000000000000000000000000000000000000000..6992ef5ea517f2a5ab3bdc9af804f364c8ea5f29
--- /dev/null
+++ b/assets/i18n/languages/fa_FA.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "ابزار شبیه سازی صدا نهایی، با دقت برای قدرت بی نظیر، مدولار بودن و تجربه کاربر پسند بهینه شده است.",
+ "This section contains some extra utilities that often may be in experimental phases.": "این بخش شامل برخی از ابزارهای اضافی است که اغلب ممکن است در مراحل ازمایشی باشد.",
+ "Output Information": "اطلاعات خروجی",
+ "The output information will be displayed here.": "اطلاعات خروجی در اینجا نمایش داده خواهد شد.",
+ "Inference": "استنباط",
+ "Train": "قطار",
+ "Extra": "اضافی",
+ "Merge Audios": "ادغام Audios",
+ "Processing": "پردازش",
+ "Audio Analyzer": "انالایزر صوتی",
+ "Model Information": "اطلاعات مدل",
+ "Plugins": "پلاگین",
+ "Download": "دانلود",
+ "Report a Bug": "گزارش یک باگ",
+ "Settings": "تنظیمات",
+ "Preprocess": "پیش پردازش",
+ "Model Name": "نام مدل",
+ "Name of the new model.": "نام مدل جدید",
+ "Enter model name": "نام مدل را وارد کنید",
+ "Dataset Path": "مسیر مجموعه داده",
+ "Path to the dataset folder.": "مسیر به پوشه مجموعه داده ها.",
+ "Refresh Datasets": "بازاوری مجموعه داده ها",
+ "Dataset Creator": "سازنده مجموعه داده",
+ "Dataset Name": "نام مجموعه داده",
+ "Name of the new dataset.": "نام مجموعه داده های جدید",
+ "Enter dataset name": "وارد کردن نام مجموعه داده ها",
+ "Upload Audio Dataset": "اپلود مجموعه داده های صوتی",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "فایل صوتی با موفقیت به مجموعه داده ها اضافه شده است. لطفا دکمه پیش پردازش را فشار دهید.",
+ "Enter dataset path": "وارد کردن مسیر مجموعه داده ها",
+ "Sampling Rate": "نرخ نمونه برداری",
+ "The sampling rate of the audio files.": "نرخ نمونه برداری از فایل های صوتی.",
+ "RVC Version": "نسخه RVC",
+ "The RVC version of the model.": "نسخه RVC از مدل.",
+ "Preprocess Dataset": "مجموعه داده پیش پردازش",
+ "Extract": "عصاره",
+ "Hop Length": "طول هاپ",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "نشان دهنده مدت زمان لازم برای سیستم برای انتقال به یک تغییر قابل توجه است. طول هاپ کوچکتر نیاز به زمان بیشتری برای استنتاج دارد اما تمایل به عملکرد دقت بالاتر دارد.",
+ "Batch Size": "اندازه دسته",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "توصیه می شود ان را با VRAM موجود GPU خود هماهنگ کنید. تنظیمات 4 ارائه می دهد دقت بهبود یافته اما پردازش کندتر، در حالی که 8 نتایج سریع تر و استاندارد را فراهم می کند.",
+ "Save Every Epoch": "ذخیره هر عصر",
+ "Determine at how many epochs the model will saved at.": "تعیین کنید که مدل در چند دوره ذخیره خواهد شد.",
+ "Total Epoch": "کل اپک",
+ "Specifies the overall quantity of epochs for the model training process.": "مقدار کلی دوره ها را برای فرایند اموزش مدل مشخص می کند.",
+ "Pretrained": "پیش اموزش دیده",
+ "Save Only Latest": "ذخیره فقط اخرین",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "فعال کردن این تنظیم منجر به صرفه جویی در فایل های G و D تنها نسخه های اخیر انها می شود و به طور موثر فضای ذخیره سازی را حفظ می کند.",
+ "Save Every Weights": "صرفه جویی در هر وزن",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "این تنظیم شما را قادر به صرفه جویی در وزن مدل در پایان هر دوره می کند.",
+ "Custom Pretrained": "سفارشی پیش اموزش دیده",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "استفاده از مدل های از پیش اموزش دیده سفارشی می تواند منجر به نتایج برتر شود، زیرا انتخاب مناسب ترین مدل های از پیش اموزش دیده متناسب با مورد استفاده خاص می تواند به طور قابل توجهی عملکرد را افزایش دهد.",
+ "Upload Pretrained Model": "اپلود مدل از پیش اموزش دیده",
+ "Refresh Custom Pretraineds": "تازه کردن Pretraineds سفارشی",
+ "Pretrained Custom Settings": "تنظیمات سفارشی از پیش اموزش داده شده",
+ "The file you dropped is not a valid pretrained file. Please try again.": "پرونده ای که حذف کرده اید یک پرونده از پیش اموزش داده شده معتبر نیست. لطفا دوباره تلاش کنید.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "با کلیک بر روی دکمه تازه کردن برای دیدن فایل از پیش اموزش دیده در منوی کشویی.",
+ "Pretrained G Path": "سفارشی پیش اموزش G",
+ "Pretrained D Path": "سفارشی از پیش اموزش دیده D",
+ "GPU Settings": "تنظیمات GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "تنظیمات پیشرفته GPU را تنظیم می کند که برای کاربران با معماری GPU بهتر توصیه می شود.",
+ "GPU Custom Settings": "تنظیمات سفارشی GPU",
+ "GPU Number": "شماره GPU",
+ "0 to ∞ separated by -": "0 به ∞ جدا شده توسط -",
+ "GPU Information": "اطلاعات GPU",
+ "Pitch Guidance": "راهنمای زمین",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "با استفاده از راهنمایی زمین، امکان پذیر است که زیر و بمی صدا اصلی، از جمله زمین ان را منعکس کند. این ویژگی به ویژه برای اواز خواندن و سناریوهای دیگر که در ان حفظ ملودی اصلی یا الگوی زمین ضروری است، ارزشمند است.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "از مدل های از پیش اموزش دیده در هنگام اموزش خود استفاده کنید. این رویکرد مدت زمان اموزش را کاهش می دهد و کیفیت کلی را افزایش می دهد.",
+ "Extract Features": "استخراج ویژگی ها",
+ "Start Training": "شروع اموزش",
+ "Generate Index": "ایجاد نمایه",
+ "Voice Model": "مدل صوتی",
+ "Select the voice model to use for the conversion.": "مدل صوتی مورد استفاده برای تبدیل را انتخاب کنید.",
+ "Index File": "پروندۀ نمایه",
+ "Select the index file to use for the conversion.": "انتخاب فایل شاخص برای استفاده برای تبدیل.",
+ "Refresh": "نوسازی",
+ "Unload Voice": "بارگیری صدا",
+ "Single": "تک",
+ "Upload Audio": "بارگذاری صدا",
+ "Select Audio": "انتخاب صدا",
+ "Select the audio to convert.": "صدا را برای تبدیل انتخاب کنید.",
+ "Advanced Settings": "تنظیمات پیشرفته",
+ "Clear Outputs (Deletes all audios in assets/audios)": "پاک کردن خروجی ها (حذف تمام فایل های صوتی در دارایی ها / audios)",
+ "Custom Output Path": "مسیر خروجی سفارشی",
+ "Output Path": "مسیر خروجی",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "مسیری که در ان صدای خروجی ذخیره می شود، به طور پیش فرض در assets/audios/output.wav",
+ "Split Audio": "تقسیم صوتی",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "صدا را به تکه های تقسیم کنید تا استنتاج شود تا در برخی موارد نتایج بهتری کسب شود.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "درخواست autotune نرم به استنتاج خود را، توصیه می شود برای تبدیل اواز.",
+ "Clean Audio": "پاک صوتی",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "خروجی صوتی خود را با استفاده از الگوریتم های تشخیص سر و صدا تمیز کنید، توصیه شده برای صحبت کردن صوتی.",
+ "Clean Strength": "قدرت پاک",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "سطح پاکسازی را به صوتی که می خواهید تنظیم کنید، هرچه بیشتر ان را افزایش دهید، بیشتر تمیز می شود، اما ممکن است صدا فشرده تر شود.",
+ "Pitch": "زمین",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "زمین صدا را تنظیم کنید، هر چه مقدار بالاتر باشد، زمین بالاتر است.",
+ "Filter Radius": "شعاع پالا",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "اگر عدد بزرگتر یا مساوی با سه باشد، استفاده از فیلتر متوسط در نتایج تن جمع اوری شده پتانسیل کاهش تنفس را دارد.",
+ "Search Feature Ratio": "نسبت ویژگی جستجو",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "تاثیر اعمال شده توسط فایل شاخص؛ ارزش بالاتر مربوط به نفوذ بیشتر است. با این حال، انتخاب مقادیر پایین تر می تواند به کاهش مصنوعات موجود در صدا کمک کند.",
+ "Volume Envelope": "پاکت جلد",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "جایگزین یا ترکیب با پاکت حجم خروجی. هر چه نسبت به 1 نزدیک تر باشد، پاکت خروجی بیشتر استفاده می شود.",
+ "Protect Voiceless Consonants": "محافظت از صامت های بی صدا",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "حفاظت از صامت های متمایز و صداهای تنفسی برای جلوگیری از پارگی الکترو اکوستیک و سایر مصنوعات. کشیدن پارامتر به حداکثر مقدار ان از 0.5 ارائه می دهد حفاظت جامع. با این حال، کاهش این مقدار ممکن است میزان حفاظت را کاهش دهد در حالی که به طور بالقوه اثر نمایه سازی را کاهش می دهد.",
+ "Pitch extraction algorithm": "الگوریتم استخراج زمین",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "الگوریتم استخراج زمین برای استفاده برای تبدیل صوتی. الگوریتم پیش فرض rmvpe است که برای اکثر موارد توصیه می شود.",
+ "Convert": "تبدیل",
+ "Export Audio": "صادرات صدا",
+ "Batch": "دسته",
+ "Input Folder": "پوشه ورودی",
+ "Select the folder containing the audios to convert.": "پوشه حاوی فایل های صوتی را برای تبدیل انتخاب کنید.",
+ "Enter input path": "وارد کردن مسیر ورودی",
+ "Output Folder": "پوشۀ خروجی",
+ "Select the folder where the output audios will be saved.": "پوشه ای را انتخاب کنید که صدای خروجی در ان ذخیره می شود.",
+ "Enter output path": "وارد کردن مسیر خروجی",
+ "Get information about the audio": "دریافت اطلاعات در مورد صدا",
+ "Information about the audio file": "اطلاعات مربوط به فایل صوتی",
+ "Waiting for information...": "در انتظار اطلاعات...",
+ "## Voice Blender": "## بلندر صوتی",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "دو مدل صوتی را انتخاب کنید، درصد ترکیب مورد نظر خود را تنظیم کنید و انها را به یک صدای کاملا جدید ترکیب کنید.",
+ "Voice Blender": "بلندر صوتی",
+ "Drag and drop your model here": "کشیدن و رها کردن مدل خود را در اینجا",
+ "You can also use a custom path.": "شما همچنین می توانید از یک مسیر سفارشی استفاده کنید.",
+ "Blend Ratio": "نسبت مخلوط",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "تنظیم موقعیت بیشتر به سمت یک طرف یا طرف دیگر، مدل را بیشتر شبیه به اول یا دوم می کند.",
+ "Fusion": "همجوشی",
+ "Path to Model": "مسیر به مدل",
+ "Enter path to model": "وارد کردن مسیر به مدل",
+ "Model information to be placed": "اطلاعات مدل قرار داده می شود",
+ "Inroduce the model information": "Inroduce اطلاعات مدل",
+ "The information to be placed in the model (You can leave it blank or put anything).": "اطلاعاتی که باید در مدل قرار داده شود (شما می توانید ان را خالی بگذارید یا هر چیزی را قرار دهید).",
+ "View model information": "مشاهده اطلاعات مدل",
+ "Introduce the model pth path": "معرفی مسیر PTH مدل",
+ "View": "مشاهده",
+ "Model extraction": "استخراج مدل",
+ "Model conversion": "تبدیل مدل",
+ "Pth file": "پرونده Pth",
+ "Output of the pth file": "خروجی پروندۀ pth",
+ "# How to Report an Issue on GitHub": "# چگونه یک مشکل را در GitHub گزارش دهیم",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. با کلیک بر روی 'ضبط صفحه نمایش' دکمه زیر برای شروع ضبط مسئله شما در حال تجربه.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. هنگامی که ضبط مسئله را به پایان رساندید، بر روی دکمه \"توقف ضبط\" کلیک کنید (همان دکمه، اما برچسب بسته به اینکه ایا شما به طور فعال ضبط می کنید یا نه تغییر می کند).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. به [GitHub Issues] (https://github.com/IAHispano/Applio/issues) بروید و بر روی دکمه \"New Issue\" کلیک کنید.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. تکمیل قالب موضوع ارائه شده، اطمینان حاصل شود که شامل جزئیات در صورت نیاز، و استفاده از بخش دارایی برای اپلود فایل ضبط شده از مرحله قبلی.",
+ "Record Screen": "صفحه ضبط",
+ "Record": "رکورد",
+ "Stop Recording": "توقف ضبط",
+ "Introduce the model .pth path": "معرفی مسیر .pth مدل",
+ "See Model Information": "مشاهده اطلاعات مدل",
+ "## Download Model": "## دانلود مدل",
+ "Model Link": "پیوند مدل",
+ "Introduce the model link": "معرفی لینک مدل",
+ "Download Model": "دانلود مدل",
+ "## Drop files": "## رها کردن پروندهها",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "فایل .pth و .index خود را به این فضا بکشید. یکی را بکشید و سپس دیگری را بکشید.",
+ "TTS Voices": "TTS صداها",
+ "Select the TTS voice to use for the conversion.": "صدای TTS را برای استفاده برای تبدیل انتخاب کنید.",
+ "Text to Synthesize": "متن برای سنتز",
+ "Enter the text to synthesize.": "متن را برای ترکیب وارد کنید.",
+ "Or you can upload a .txt file": "یا شما می توانید یک فایل .txt اپلود کنید",
+ "Enter text to synthesize": "متن را برای سنتز وارد کنید",
+ "Output Path for TTS Audio": "مسیر خروجی برای صدای TTS",
+ "Output Path for RVC Audio": "مسیر خروجی برای RVC Audio",
+ "Enable Applio integration with Discord presence": "ادغام Applio را با حضور Discord فعال کنید",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "این امکان نمایش فعالیت Applio فعلی در Discord را فعال خواهد کرد.",
+ "Enable Applio integration with applio.org/models using flask": "فعال کردن ادغام Applio با applio.org/models با استفاده از فلاسک",
+ "It will activate the possibility of downloading models with a click from the website.": "این امکان دانلود مدل ها را با یک کلیک از وب سایت فعال می کند.",
+ "Theme": "تم",
+ "Select the theme you want to use. (Requires restarting Applio)": "زمینه ای را که می خواهید استفاده کنید انتخاب کنید. (نیاز به راه اندازی مجدد Applio)",
+ "Language": "زبان",
+ "Select the language you want to use. (Requires restarting Applio)": "زبانی را که می خواهید استفاده کنید انتخاب کنید. (نیاز به راه اندازی مجدد Applio)",
+ "Plugin Installer": "نصب افزونه",
+ "Drag your plugin.zip to install it": "plugin.zip خود را بکشید تا ان را نصب کنید",
+ "Version Checker": "بررسی نسخه",
+ "Check which version of Applio is the latest to see if you need to update.": "بررسی کنید که کدام نسخه از Applio اخرین است تا ببینید ایا شما نیاز به به روز رسانی دارید.",
+ "Check for updates": "بررسی برای به روز رسانی"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/fr_FR.json b/assets/i18n/languages/fr_FR.json
new file mode 100644
index 0000000000000000000000000000000000000000..a990d855a9a32764ef20145414a025837616049a
--- /dev/null
+++ b/assets/i18n/languages/fr_FR.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Outil ultime de clonage vocal, méticuleusement optimisé pour une puissance, une modularité et une expérience conviviales inégalées.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Cette section contient quelques utilitaires supplémentaires qui peuvent souvent être en phase expérimentale.",
+ "Output Information": "Informations de sortie",
+ "The output information will be displayed here.": "Les informations de sortie seront affichées ici.",
+ "Inference": "Inférence",
+ "Train": "Train",
+ "Extra": "Supplémentaire",
+ "Merge Audios": "Fusionner les audios",
+ "Processing": "Traitement",
+ "Audio Analyzer": "Analyseur audio",
+ "Model Information": "Informations sur le modèle",
+ "Plugins": "Plugins (Plugins)",
+ "Download": "Télécharger",
+ "Report a Bug": "Signaler un bogue",
+ "Settings": "Paramètres",
+ "Preprocess": "Pré-traitement",
+ "Model Name": "Nom du modèle",
+ "Name of the new model.": "Nom du nouveau modèle.",
+ "Enter model name": "Entrez le nom du modèle",
+ "Dataset Path": "Chemin d’accès au jeu de données",
+ "Path to the dataset folder.": "Chemin d’accès au dossier du jeu de données.",
+ "Refresh Datasets": "Actualiser les jeux de données",
+ "Dataset Creator": "Créateur de jeux de données",
+ "Dataset Name": "Nom du jeu de données",
+ "Name of the new dataset.": "Nom du nouveau jeu de données.",
+ "Enter dataset name": "Entrez le nom du jeu de données",
+ "Upload Audio Dataset": "Télécharger un jeu de données audio",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Le fichier audio a été ajouté avec succès au jeu de données. Veuillez cliquer sur le bouton de prétraitement.",
+ "Enter dataset path": "Entrez le chemin d’accès au jeu de données",
+ "Sampling Rate": "Fréquence d’échantillonnage",
+ "The sampling rate of the audio files.": "Fréquence d’échantillonnage des fichiers audio.",
+ "RVC Version": "RVC Version",
+ "The RVC version of the model.": "La version RVC du modèle.",
+ "Preprocess Dataset": "Jeu de données de prétraitement",
+ "Extract": "Extraire",
+ "Hop Length": "Longueur du houblon",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Indique le temps qu’il faut au système pour passer à un changement de hauteur significatif. Les longueurs de saut plus petites nécessitent plus de temps pour l’inférence, mais ont tendance à donner une plus grande précision de hauteur.",
+ "Batch Size": "Taille du lot",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Il est conseillé de l’aligner sur la VRAM disponible de votre GPU. Un réglage de 4 offre une précision améliorée mais un traitement plus lent, tandis que 8 fournit des résultats plus rapides et standard.",
+ "Save Every Epoch": "Sauvez toutes les époques",
+ "Determine at how many epochs the model will saved at.": "Déterminez à combien d’époques le modèle sera enregistré.",
+ "Total Epoch": "Époque totale",
+ "Specifies the overall quantity of epochs for the model training process.": "Spécifie la quantité globale d’époques pour le processus d’entraînement du modèle.",
+ "Pretrained": "Pré-entraîné",
+ "Save Only Latest": "Enregistrer uniquement les plus récents",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "L’activation de ce paramètre permet aux fichiers G et D d’enregistrer uniquement leurs versions les plus récentes, ce qui permet d’économiser efficacement de l’espace de stockage.",
+ "Save Every Weights": "Économisez tous les poids",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Ce paramètre vous permet d’enregistrer les poids du modèle à la fin de chaque époque.",
+ "Custom Pretrained": "Pré-entraîné sur mesure",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "L’utilisation de modèles préentraînés personnalisés peut conduire à des résultats supérieurs, car la sélection des modèles préentraînés les plus appropriés et adaptés au cas d’utilisation spécifique peut améliorer considérablement les performances.",
+ "Upload Pretrained Model": "Télécharger un modèle pré-entraîné",
+ "Refresh Custom Pretraineds": "Actualiser les pré-entraînés personnalisés",
+ "Pretrained Custom Settings": "Paramètres personnalisés préentraînés",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Le fichier que vous avez déposé n’est pas un fichier pré-entraîné valide. Veuillez réessayer.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Cliquez sur le bouton d’actualisation pour afficher le fichier préentraîné dans le menu déroulant.",
+ "Pretrained G Path": "G pré-entraîné personnalisé",
+ "Pretrained D Path": "D pré-entraîné personnalisé",
+ "GPU Settings": "Paramètres GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Définit les paramètres GPU avancés, recommandés pour les utilisateurs disposant d’une meilleure architecture GPU.",
+ "GPU Custom Settings": "Paramètres personnalisés du GPU",
+ "GPU Number": "Numéro de GPU",
+ "0 to ∞ separated by -": "0 à ∞ séparés par -",
+ "GPU Information": "Informations sur le GPU",
+ "Pitch Guidance": "Guidage de la hauteur",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "En utilisant le guidage de hauteur, il devient possible de refléter l’intonation de la voix d’origine, y compris sa hauteur. Cette fonctionnalité est particulièrement utile pour le chant et d’autres scénarios où la préservation de la mélodie ou du modèle de hauteur d’origine est essentielle.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Utilisez des modèles pré-entraînés lors de l’entraînement des vôtres. Cette approche permet de réduire la durée de la formation et d’améliorer la qualité globale.",
+ "Extract Features": "Extraire des caractéristiques",
+ "Start Training": "Commencer l’entraînement",
+ "Generate Index": "Générer un index",
+ "Voice Model": "Modèle vocal",
+ "Select the voice model to use for the conversion.": "Sélectionnez le modèle vocal à utiliser pour la conversion.",
+ "Index File": "Fichier d’index",
+ "Select the index file to use for the conversion.": "Sélectionnez le fichier d’index à utiliser pour la conversion.",
+ "Refresh": "Rafraîchir",
+ "Unload Voice": "Décharger la voix",
+ "Single": "Célibataire",
+ "Upload Audio": "Télécharger l’audio",
+ "Select Audio": "Sélectionnez Audio (Audio)",
+ "Select the audio to convert.": "Sélectionnez l’audio à convertir.",
+ "Advanced Settings": "Paramètres avancés",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Effacer les sorties (supprime tous les audios dans les ressources/audios)",
+ "Custom Output Path": "Chemin de sortie personnalisé",
+ "Output Path": "Chemin de sortie",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Le chemin où l’audio de sortie sera enregistré, par défaut dans assets/audios/output.wav",
+ "Split Audio": "Diviser l’audio",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Divisez l’audio en morceaux pour l’inférence afin d’obtenir de meilleurs résultats dans certains cas.",
+ "Autotune": "Réglage automatique",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Appliquez un réglage automatique doux à vos inférences, recommandé pour les conversions de chant.",
+ "Clean Audio": "Audio clair",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Nettoyez votre sortie audio à l’aide d’algorithmes de détection de bruit, recommandés pour les audios parlants.",
+ "Clean Strength": "Force propre",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Réglez le niveau de nettoyage sur l’audio que vous souhaitez, plus vous l’augmentez plus il nettoiera, mais il est possible que l’audio soit plus compressé.",
+ "Pitch": "Tanguer",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Réglez la hauteur de l’audio, plus la valeur est élevée, plus la hauteur est élevée.",
+ "Filter Radius": "Rayon du filtre",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Si le nombre est supérieur ou égal à trois, l’utilisation d’un filtrage médian sur les résultats de tonalité recueillis a le potentiel de diminuer la respiration.",
+ "Search Feature Ratio": "Rapport de caractéristiques de recherche",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Influence exercée par le fichier d’index ; Une valeur plus élevée correspond à une plus grande influence. Cependant, opter pour des valeurs plus faibles peut aider à atténuer les artefacts présents dans l’audio.",
+ "Volume Envelope": "Enveloppe de volume",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Remplacez ou fusionnez avec l’enveloppe de volume de la sortie. Plus le rapport est proche de 1, plus l’enveloppe de sortie est utilisée.",
+ "Protect Voiceless Consonants": "Protéger les consonnes sourdes",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Protégez les consonnes distinctes et les sons respiratoires pour éviter les déchirures électroacoustiques et autres artefacts. L’extraction du paramètre à sa valeur maximale de 0,5 offre une protection complète. Toutefois, la réduction de cette valeur peut réduire l’étendue de la protection tout en atténuant potentiellement l’effet d’indexation.",
+ "Pitch extraction algorithm": "Algorithme d’extraction de hauteur",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Algorithme d’extraction de hauteur à utiliser pour la conversion audio. L’algorithme par défaut est rmvpe, qui est recommandé dans la plupart des cas.",
+ "Convert": "Convertir",
+ "Export Audio": "Exporter l’audio",
+ "Batch": "Lot",
+ "Input Folder": "Dossier d’entrée",
+ "Select the folder containing the audios to convert.": "Sélectionnez le dossier contenant les audios à convertir.",
+ "Enter input path": "Entrez le chemin d’entrée",
+ "Output Folder": "Dossier de sortie",
+ "Select the folder where the output audios will be saved.": "Sélectionnez le dossier dans lequel les audios de sortie seront enregistrés.",
+ "Enter output path": "Entrez le chemin de sortie",
+ "Get information about the audio": "Obtenir des informations sur l’audio",
+ "Information about the audio file": "Informations sur le fichier audio",
+ "Waiting for information...": "En attente d’informations...",
+ "## Voice Blender": "## Mélangeur de voix",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Sélectionnez deux modèles de voix, définissez le pourcentage de fusion souhaité et mélangez-les en une toute nouvelle voix.",
+ "Voice Blender": "Mélangeur de voix",
+ "Drag and drop your model here": "Faites glisser et déposez votre modèle ici",
+ "You can also use a custom path.": "Vous pouvez également utiliser un chemin personnalisé.",
+ "Blend Ratio": "Rapport de mélange",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "En ajustant la position d’un côté ou de l’autre, le modèle sera plus similaire au premier ou au second.",
+ "Fusion": "Fusion",
+ "Path to Model": "Chemin d’accès au modèle",
+ "Enter path to model": "Entrez le chemin d’accès au modèle",
+ "Model information to be placed": "Informations sur le modèle à placer",
+ "Inroduce the model information": "Introduire les informations du modèle",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Les informations à placer dans le modèle (vous pouvez les laisser vides ou mettre n’importe quoi).",
+ "View model information": "Afficher les informations sur le modèle",
+ "Introduce the model pth path": "Présentation du modèle pth path",
+ "View": "Vue",
+ "Model extraction": "Extraction de modèles",
+ "Model conversion": "Conversion de modèle",
+ "Pth file": "Fichier Pth",
+ "Output of the pth file": "Sortie du fichier pth",
+ "# How to Report an Issue on GitHub": "# Comment signaler un problème sur GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Cliquez sur le bouton « Enregistrer l’écran » ci-dessous pour commencer à enregistrer le problème que vous rencontrez.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Une fois que vous avez terminé d’enregistrer le numéro, cliquez sur le bouton « Arrêter l’enregistrement » (le même bouton, mais l’étiquette change selon que vous enregistrez activement ou non).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Allez dans [GitHub Issues](https://github.com/IAHispano/Applio/issues) et cliquez sur le bouton « Nouveau problème ».",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Complétez le modèle de problème fourni, en veillant à inclure les détails nécessaires, et utilisez la section des ressources pour télécharger le fichier enregistré de l’étape précédente.",
+ "Record Screen": "Écran d’enregistrement",
+ "Record": "Enregistrer",
+ "Stop Recording": "Arrêter l’enregistrement",
+ "Introduce the model .pth path": "Présentation du chemin d’accès .pth du modèle",
+ "See Model Information": "Voir les informations sur le modèle",
+ "## Download Model": "## Télécharger le modèle",
+ "Model Link": "Lien vers le modèle",
+ "Introduce the model link": "Introduire le lien du modèle",
+ "Download Model": "Télécharger le modèle",
+ "## Drop files": "## Déposer des fichiers",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Faites glisser vos fichiers .pth et .index dans cet espace. Faites glisser l’un, puis l’autre.",
+ "TTS Voices": "Voix TTS",
+ "Select the TTS voice to use for the conversion.": "Sélectionnez la voix TTS à utiliser pour la conversion.",
+ "Text to Synthesize": "Texte à synthétiser",
+ "Enter the text to synthesize.": "Saisissez le texte à synthétiser.",
+ "Or you can upload a .txt file": "Vous pouvez également télécharger un fichier .txt",
+ "Enter text to synthesize": "Saisir le texte à synthétiser",
+ "Output Path for TTS Audio": "Chemin de sortie pour l’audio TTS",
+ "Output Path for RVC Audio": "Chemin de sortie pour l’audio RVC",
+ "Enable Applio integration with Discord presence": "Activer l’intégration d’Applio avec la présence de Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Il activera la possibilité d’afficher l’activité actuelle d’Applio dans Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Activer l’intégration d’Applio avec applio.org/models à l’aide de flask",
+ "It will activate the possibility of downloading models with a click from the website.": "Il activera la possibilité de télécharger des modèles en un clic depuis le site web.",
+ "Theme": "Thème",
+ "Select the theme you want to use. (Requires restarting Applio)": "Sélectionnez le thème que vous souhaitez utiliser. (Nécessite le redémarrage d’Applio)",
+ "Language": "Langue",
+ "Select the language you want to use. (Requires restarting Applio)": "Sélectionnez la langue que vous souhaitez utiliser. (Nécessite le redémarrage d’Applio)",
+ "Plugin Installer": "Installateur de plugin",
+ "Drag your plugin.zip to install it": "Faites glisser votre plugin.zip pour l’installer",
+ "Version Checker": "Vérificateur de version",
+ "Check which version of Applio is the latest to see if you need to update.": "Vérifiez quelle version d’Applio est la plus récente pour voir si vous devez effectuer une mise à jour.",
+ "Check for updates": "Vérifier les mises à jour"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/gu_GU.json b/assets/i18n/languages/gu_GU.json
new file mode 100644
index 0000000000000000000000000000000000000000..27b48a084d5ef5cd8e39b40ee0f0471813afa2a1
--- /dev/null
+++ b/assets/i18n/languages/gu_GU.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "અંતિમ વોઇસ ક્લોનિંગ ટૂલ, જે અજોડ શક્તિ, મોડ્યુલારિટી અને વપરાશકર્તા-મૈત્રીપૂર્ણ અનુભવ માટે સાવચેતીપૂર્વક ઓપ્ટિમાઇઝ કરવામાં આવ્યું છે.",
+ "This section contains some extra utilities that often may be in experimental phases.": "આ વિભાગમાં કેટલીક વધારાની ઉપયોગિતાઓ છે જે ઘણીવાર પ્રાયોગિક તબક્કામાં હોઈ શકે છે.",
+ "Output Information": "આઉટપુટ જાણકારી",
+ "The output information will be displayed here.": "આઉટપુટ જાણકારી અંહિ દર્શાવવામાં આવશે.",
+ "Inference": "અનુમાન",
+ "Train": "રેલગાડી",
+ "Extra": "વધારાનું",
+ "Merge Audios": "ઓડિયો ભેગા કરો",
+ "Processing": "પ્રક્રિયા કરી રહ્યા છીએ",
+ "Audio Analyzer": "ઓડિયો વિશ્લેષક",
+ "Model Information": "મોડેલ જાણકારી",
+ "Plugins": "પ્લગઇનો",
+ "Download": "ડાઉનલોડ",
+ "Report a Bug": "ભૂલનો અહેવાલ આપો",
+ "Settings": "સુયોજનો",
+ "Preprocess": "પ્રીપ્રોસેસ",
+ "Model Name": "મોડેલ નામ",
+ "Name of the new model.": "નવા મોડેલનું નામ.",
+ "Enter model name": "મોડેલ નામ દાખલ કરો",
+ "Dataset Path": "ડેટાસેટ પાથ",
+ "Path to the dataset folder.": "ડેટાસેટ ફોલ્ડરનો પાથ.",
+ "Refresh Datasets": "ડેટાસેટ્સ પુનઃતાજું કરો",
+ "Dataset Creator": "ડેટાસેટ બનાવનાર",
+ "Dataset Name": "ડેટાસેટ નામ",
+ "Name of the new dataset.": "નવા ડેટાસેટનું નામ.",
+ "Enter dataset name": "ડેટાસેટ નામ દાખલ કરો",
+ "Upload Audio Dataset": "ઓડિયો ડેટાસેટ અપલોડ કરો",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ઓડિયો ફાઈલ સફળતાપૂર્વક ડેટાસેટમાં ઉમેરાઈ ગઈ છે. કૃપા કરીને પ્રીપ્રોસેસ બટન પર ક્લિક કરો.",
+ "Enter dataset path": "ડેટાસેટ પાથને દાખલ કરો",
+ "Sampling Rate": "નમૂનાનો દર",
+ "The sampling rate of the audio files.": "ઓડિયો ફાઈલોનો નમૂનાનો દર.",
+ "RVC Version": "RVC આવૃત્તિ",
+ "The RVC version of the model.": "મોડેલનું આરવીસી વર્ઝન.",
+ "Preprocess Dataset": "પ્રીપ્રોસેસ ડેટાસેટ",
+ "Extract": "અર્ક કાઢો",
+ "Hop Length": "હોપ લંબાઈ",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "સિસ્ટમને નોંધપાત્ર પિચ પરિવર્તનમાં સંક્રમણ કરવામાં જે સમયગાળો લાગે છે તે સૂચવે છે. નાના હોપની લંબાઈને અનુમાન માટે વધુ સમયની જરૂર હોય છે પરંતુ ઉચ્ચ પિચ ચોકસાઈ પ્રાપ્ત કરવાનું વલણ ધરાવે છે.",
+ "Batch Size": "બેચ માપ",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "તેને તમારા જીપીયુના ઉપલબ્ધ વીઆરએએમ સાથે સંરેખિત કરવાની સલાહ આપવામાં આવે છે. 4નું સેટિંગ સુધારેલી ચોકસાઈ આપે છે પરંતુ પ્રક્રિયા ધીમી છે, જ્યારે 8 ઝડપી અને પ્રમાણભૂત પરિણામો પ્રદાન કરે છે.",
+ "Save Every Epoch": "દરેક ઈપોકનો સંગ્રહ કરો",
+ "Determine at how many epochs the model will saved at.": "મોડેલ કેટલા યુગમાં સંગ્રહાશે તે નક્કી કરો.",
+ "Total Epoch": "કુલ ઈપોક",
+ "Specifies the overall quantity of epochs for the model training process.": "મોડેલ તાલીમ પ્રક્રિયા માટે યુગોનો એકંદર જથ્થો સ્પષ્ટ કરે છે.",
+ "Pretrained": "પૂર્વપ્રશિક્ષિત",
+ "Save Only Latest": "ફક્ત તાજેતરનાં ને સંગ્રહો",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "આ ગોઠવણને સક્રિય કરવાથી જી અને ડી (G) અને ડી (D) ફાઇલો માત્ર તેમના તાજેતરના વર્ઝનને જ સેવ કરી શકશે, અસરકારક રીતે સ્ટોરેજ સ્પેસનું સંરક્ષણ કરશે.",
+ "Save Every Weights": "દરેક વજનોને બચાવો",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "આ ગોઠવણી તમને દરેક યુગના અંતે મોડેલના વજનને બચાવવા માટે સક્ષમ બનાવે છે.",
+ "Custom Pretrained": "વૈવિધ્યપૂર્ણ પૂર્વટ્રેઈન થયેલ",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "કસ્ટમ પ્રીટ્રેઇન્ડ મોડેલ્સનો ઉપયોગ કરવાથી શ્રેષ્ઠ પરિણામો મળી શકે છે, કારણ કે ચોક્કસ ઉપયોગના કિસ્સાને અનુરૂપ સૌથી યોગ્ય પ્રિટ્રેઇન્ડ મોડેલ્સની પસંદગી કરવાથી કામગીરીમાં નોંધપાત્ર વધારો થઈ શકે છે.",
+ "Upload Pretrained Model": "પહેલેથી પ્રશિક્ષિત મોડેલ અપલોડ કરો",
+ "Refresh Custom Pretraineds": "કસ્ટમ પૂર્વપ્રશિક્ષિતોને તાજુ કરો",
+ "Pretrained Custom Settings": "પહેલેથી પ્રશિક્ષિત વૈવિધ્યપૂર્ણ સુયોજનો",
+ "The file you dropped is not a valid pretrained file. Please try again.": "તમે મૂકેલી ફાઇલ એ યોગ્ય પૂર્વતાલીમવાળી ફાઇલ નથી. કૃપા કરીને ફરી પ્રયાસ કરો.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "ડ્રોપડાઉન મેનુમાં પહેલેથી તાલીમ લીધેલ ફાઈલ જોવા માટે રિફ્રેશ બટન પર ક્લિક કરો.",
+ "Pretrained G Path": "વૈવિધ્યપૂર્ણ પૂર્વપ્રશિક્ષિત G",
+ "Pretrained D Path": "વૈવિધ્યપૂર્ણ પૂર્વપ્રશિક્ષિત D",
+ "GPU Settings": "GPU સુયોજનો",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "અદ્યતન GPU ગોઠવણો સુયોજિત કરે છે, જે વધુ સારા GPU આર્કિટેક્ચર ધરાવતા વપરાશકર્તાઓ માટે આગ્રહણીય છે.",
+ "GPU Custom Settings": "GPU કસ્ટમ સુયોજનો",
+ "GPU Number": "GPU નંબર",
+ "0 to ∞ separated by -": "0 થી ∞ આના દ્વારા અલગ પાડવામાં આવે છે -",
+ "GPU Information": "GPU જાણકારી",
+ "Pitch Guidance": "પિચ માર્ગદર્શન",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "પિચ માર્ગદર્શનનો ઉપયોગ કરીને, મૂળ અવાજના રણકારને પ્રતિબિંબિત કરવાનું શક્ય બને છે, જેમાં તેની પીચનો પણ સમાવેશ થાય છે. આ સુવિધા ખાસ કરીને ગાવા અને અન્ય દૃશ્યો માટે મૂલ્યવાન છે જ્યાં મૂળ મેલોડી અથવા પિચ પેટર્નને સાચવવી જરૂરી છે.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "તમારા પોતાનાને તાલીમ આપતી વખતે પૂર્વપ્રશિક્ષિત મોડેલોનો ઉપયોગ કરો. આ અભિગમ તાલીમનો સમયગાળો ઘટાડે છે અને એકંદર ગુણવત્તામાં વધારો કરે છે.",
+ "Extract Features": "લક્ષણોનો અર્ક કાઢો",
+ "Start Training": "તાલીમ શરૂ કરો",
+ "Generate Index": "અનુક્રમણિકા બનાવો",
+ "Voice Model": "અવાજ મોડેલ",
+ "Select the voice model to use for the conversion.": "રૂપાંતરણ માટે વાપરવા માટે વોઇસ મોડેલ પસંદ કરો.",
+ "Index File": "અનુક્રમણિકા ફાઇલ",
+ "Select the index file to use for the conversion.": "રૂપાંતરણ માટે વાપરવા માટે અનુક્રમણિકા ફાઈલ પસંદ કરો.",
+ "Refresh": "પુનઃતાજું કરો",
+ "Unload Voice": "અવાજ અનલોડ કરો",
+ "Single": "એકલું",
+ "Upload Audio": "ઓડિયો અપલોડ કરો",
+ "Select Audio": "ઓડિયો પસંદ કરો",
+ "Select the audio to convert.": "રૂપાંતરિત કરવા માટે ઓડિયો પસંદ કરો.",
+ "Advanced Settings": "અદ્યતન સુયોજનો",
+ "Clear Outputs (Deletes all audios in assets/audios)": "આઉટપુટ સાફ કરો (સંપત્તિઓ/ઓડિયોમાં બધા ઓડિયો કાઢી નાંખે છે)",
+ "Custom Output Path": "કસ્ટમ આઉટપુટ પાથ",
+ "Output Path": "આઉટપુટ પાથ",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "પાથ કે જ્યાં આઉટપુટ ઓડિયો એ મૂળભૂત રીતે assets/audios/output.wav માં સંગ્રહાશે",
+ "Split Audio": "ઓડિયો વિભાજિત કરો",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "કેટલાક કિસ્સાઓમાં વધુ સારા પરિણામો મેળવવા માટે અનુમાન માટે ઓડિઓને ભાગોમાં વિભાજિત કરો.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "તમારા અનુમાનો પર નરમ ઓટોટ્યુન લાગુ કરો, જે ગાવા માટે ભલામણ કરવામાં આવે છે રૂપાંતરણો ગાવા માટે ભલામણ કરવામાં આવે છે.",
+ "Clean Audio": "ઓડિયો સાફ કરો",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "ઓડિયો બોલવા માટે ભલામણ કરવામાં આવેલા નોઇઝ ડિટેક્શન એલ્ગોરિધમ્સનો ઉપયોગ કરીને તમારા ઓડિયો આઉટપુટને સાફ કરો.",
+ "Clean Strength": "સ્વચ્છ મજબૂતાઈ",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "તમે ઇચ્છો તે ઓડિયો પર સફાઇનું સ્તર સુયોજિત કરો, તમે તેને જેટલું વધારશો તેટલું તે સાફ થશે, પરંતુ શક્ય છે કે ઓડિયો વધુ સંકુચિત હશે.",
+ "Pitch": "પિચ",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ઓડિયોની પિચ સુયોજિત કરો, કિંમત જેટલી ઊંચી હશે, તેટલી પીચ વધુ ઊંચી હશે.",
+ "Filter Radius": "ફિલ્ટર ત્રિજ્યા",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "જો સંખ્યા ત્રણ કરતા વધારે અથવા સમાન હોય, તો એકત્રિત સ્વર પરિણામો પર મધ્યમ ફિલ્ટરિંગનો ઉપયોગ કરવાથી શ્વસનનક્કી થવાની સંભાવના રહે છે.",
+ "Search Feature Ratio": "શોધ લક્ષણ ગુણોત્તર",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "અનુક્રમણિકા ફાઈલ દ્વારા લાગતો પ્રભાવ; ઉચ્ચતર મૂલ્ય વધુ પ્રભાવને મળતું આવે છે. જો કે, નીચા મૂલ્યોની પસંદગી કરવાથી ઓડિયોમાં હાજર કલાકૃતિઓને ઘટાડવામાં મદદ મળી શકે છે.",
+ "Volume Envelope": "વોલ્યુમ કવરpaper size",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "આઉટપુટના વોલ્યુમ પરબિડીયા સાથે અવેજી અથવા મિશ્રણ કરો. ગુણોત્તર 1 ની નજીક હોય છે, આઉટપુટ પરબિડીયાને વધુ ઉપયોગમાં લેવામાં આવે છે.",
+ "Protect Voiceless Consonants": "અવાજ વગરના વ્યંજનોનું રક્ષણ કરો",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "ઇલેક્ટ્રો-એકોસ્ટિક ફાટી ન જાય અને અન્ય કલાકૃતિઓ ન ફાટી જાય તે માટે વિશિષ્ટ વ્યંજનો અને શ્વાસોચ્છવાસના અવાજનું રક્ષણ કરો. પરિમાણને તેના ૦.૫ ના મહત્તમ મૂલ્ય તરફ ખેંચવું એ વ્યાપક સુરક્ષા પ્રદાન કરે છે. જો કે, આ મૂલ્યમાં ઘટાડો કરવાથી અનુક્રમણિકાની અસરને સંભવિતપણે ઘટાડવાની સાથે સંરક્ષણની હદમાં ઘટાડો થઈ શકે છે.",
+ "Pitch extraction algorithm": "પિચ નિષ્કર્ષણ અલગોરિધમ",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ઓડિઓ રૂપાંતર માટે વાપરવા માટે પિચ નિષ્કર્ષણ અલ્ગોરિધમનો. મૂળભૂત અલ્ગોરિધમ એ rmvpe છે, જે મોટાભાગના કિસ્સાઓમાં ભલામણ કરવામાં આવે છે.",
+ "Convert": "રૂપાંતર કરો",
+ "Export Audio": "ઓડિયો નિકાસ કરો",
+ "Batch": "બેચ",
+ "Input Folder": "ઇનપુટ ફોલ્ડર",
+ "Select the folder containing the audios to convert.": "ફેરવવા માટે ઓડિયો સમાવતા ફોલ્ડરને પસંદ કરો.",
+ "Enter input path": "ઇનપુટ પાથને દાખલ કરો",
+ "Output Folder": "આઉટપુટ ફોલ્ડર",
+ "Select the folder where the output audios will be saved.": "ફોલ્ડર પસંદ કરો કે જ્યાં આઉટપુટ ઓડિયો સંગ્રહાશે.",
+ "Enter output path": "આઉટપુટ પાથ દાખલ કરો",
+ "Get information about the audio": "ઓડિયો વિશે જાણકારી મેળવો",
+ "Information about the audio file": "ઓડિયો ફાઈલ વિશેની માહિતી",
+ "Waiting for information...": "જાણકારી માટે રાહ જોઇ રહ્યા છીએ...",
+ "## Voice Blender": "## વોઇસ બ્લેન્ડર",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "બે વોઇસ મોડલ્સ પસંદ કરો, તમારી ઇચ્છિત મિશ્રણ ટકાવારી સેટ કરો અને તેને સંપૂર્ણપણે નવા અવાજમાં મિશ્રિત કરો.",
+ "Voice Blender": "અવાજ બ્લેન્ડર",
+ "Drag and drop your model here": "તમારા મોડેલને અંહિ ખેંચો અને મૂકો",
+ "You can also use a custom path.": "તમે કસ્ટમ પાથનો પણ ઉપયોગ કરી શકો છો.",
+ "Blend Ratio": "મિશ્રણ ગુણોત્તર",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "સ્થિતિને એક બાજુ અથવા બીજી બાજુ વધુ સમાયોજિત કરવાથી મોડેલને પ્રથમ અથવા બીજી બાજુ વધુ સમાન બનાવશે.",
+ "Fusion": "ફ્યુઝન",
+ "Path to Model": "મોડેલનો પાથ",
+ "Enter path to model": "મોડેલ માટે પાથને દાખલ કરો",
+ "Model information to be placed": "મૂકવાની મોડેલ માહિતી",
+ "Inroduce the model information": "મોડેલની જાણકારીને ઇનરોડ્યુસ કરો",
+ "The information to be placed in the model (You can leave it blank or put anything).": "મોડેલમાં મૂકવાની માહિતી (તમે તેને ખાલી છોડી શકો છો અથવા કંઈપણ મૂકી શકો છો).",
+ "View model information": "મોડેલ જાણકારી જુઓ",
+ "Introduce the model pth path": "મોડેલ pth પાથનો પરિચય આપો",
+ "View": "જુઓ",
+ "Model extraction": "મોડેલ નિષ્કર્ષણ",
+ "Model conversion": "મોડેલ રૂપાંતરણ",
+ "Pth file": "Pth ફાઈલ",
+ "Output of the pth file": "pth ફાઇલનું આઉટપુટ",
+ "# How to Report an Issue on GitHub": "# GitHub પર કોઈ સમસ્યાની જાણ કેવી રીતે કરવી",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. તમે જે સમસ્યાનો સામનો કરી રહ્યા છો તેને રેકોર્ડ કરવાનું શરૂ કરવા માટે નીચે આપેલા 'રેકોર્ડ સ્ક્રીન' બટન પર ક્લિક કરો.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. એક વખત તમે સમસ્યાનું રેકોર્ડિંગ પૂરું કરી લો એટલે 'સ્ટોપ રેકોર્ડિંગ' બટન પર ક્લિક કરો (આ જ બટન, પરંતુ તમે સક્રિયપણે રેકોર્ડિંગ કરી રહ્યા છો કે નહીં તેના આધારે લેબલ બદલાય છે).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub issues] (https://github.com/IAHispano/Applio/issues) પર જાઓ અને 'ન્યૂ ઇશ્યૂ' બટન પર ક્લિક કરો.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. પૂરી પાડવામાં આવેલ ઇશ્યૂ ટેમ્પલેટ પૂર્ણ કરો, જરૂરિયાત મુજબ વિગતોનો સમાવેશ કરવાની ખાતરી કરો અને અગાઉના પગલામાંથી રેકોર્ડ કરેલી ફાઇલને અપલોડ કરવા માટે અસ્કયામતોના વિભાગનો ઉપયોગ કરો.",
+ "Record Screen": "રેકોર્ડ સ્ક્રીન",
+ "Record": "રેકોર્ડ",
+ "Stop Recording": "રેકોર્ડ કરવાનું બંધ કરો",
+ "Introduce the model .pth path": "મોડલ .pth પાથને રજૂ કરો",
+ "See Model Information": "મોડેલ જાણકારી જુઓ",
+ "## Download Model": "## ડાઉનલોડ મોડેલ",
+ "Model Link": "મોડેલ કડી",
+ "Introduce the model link": "મોડેલ કડીનો પરિચય આપો",
+ "Download Model": "ડાઉનલોડ મોડેલ",
+ "## Drop files": "## ફાઇલો મૂકો",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "આ જગ્યામાં તમારી .pth ફાઇલ અને .index ફાઇલને ખેંચો. એકને ખેંચો અને પછી બીજું.",
+ "TTS Voices": "TTS અવાજો",
+ "Select the TTS voice to use for the conversion.": "રૂપાંતરણ માટે વાપરવા માટે TTS અવાજ પસંદ કરો.",
+ "Text to Synthesize": "સંશ્લેષણ કરવા માટેનું લખાણ",
+ "Enter the text to synthesize.": "સંશ્લેષણ કરવા માટે લખાણ દાખલ કરો.",
+ "Or you can upload a .txt file": "અથવા તમે .txt ફાઇલ અપલોડ કરી શકો છો",
+ "Enter text to synthesize": "સંશ્લેષણ કરવા માટે લખાણ દાખલ કરો",
+ "Output Path for TTS Audio": "TTS ઓડિયો માટે આઉટપુટ પાથ",
+ "Output Path for RVC Audio": "RVC ઓડિયો માટે આઉટપુટ પાથ",
+ "Enable Applio integration with Discord presence": "ડિસ્કોર્ડ હાજરી સાથે એપ્લિઓ સંકલન સક્રિય કરો",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "તે ડિસ્કોર્ડમાં વર્તમાન એપ્લિયો પ્રવૃત્તિ પ્રદર્શિત કરવાની સંભાવનાને સક્રિય કરશે.",
+ "Enable Applio integration with applio.org/models using flask": "ફ્લાસ્કની મદદથી applio.org/models સાથે એપ્લિયો સંકલન સક્રિય કરો",
+ "It will activate the possibility of downloading models with a click from the website.": "તે વેબસાઇટ પરથી એક ક્લિક સાથે મોડેલો ડાઉનલોડ કરવાની સંભાવનાને સક્રિય કરશે.",
+ "Theme": "થીમ",
+ "Select the theme you want to use. (Requires restarting Applio)": "થીમ પસંદ કરો જે તમે વાપરવા માંગો છો. (એપ્લિયોને ફરી શરૂ કરવાની જરૂર છે)",
+ "Language": "ભાષા",
+ "Select the language you want to use. (Requires restarting Applio)": "તમે જે ભાષા વાપરવા માંગો તે પસંદ કરો. (એપ્લિયોને ફરી શરૂ કરવાની જરૂર છે)",
+ "Plugin Installer": "પ્લગઇન સ્થાપનાર",
+ "Drag your plugin.zip to install it": "તેને સ્થાપિત કરવા માટે તમારા plugin.zip ખેંચો",
+ "Version Checker": "આવૃત્તિ ચકાસનાર",
+ "Check which version of Applio is the latest to see if you need to update.": "તમારે અપડેટ કરવાની જરૂર છે કે નહીં તે જોવા માટે એપ્લિઓનું કયું સંસ્કરણ નવીનતમ છે તે તપાસો.",
+ "Check for updates": "સુધારાઓ માટે ચકાસો"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/hi_HI.json b/assets/i18n/languages/hi_HI.json
new file mode 100644
index 0000000000000000000000000000000000000000..da116a08f8177508e0d07deb3c2e1b0921544fce
--- /dev/null
+++ b/assets/i18n/languages/hi_HI.json
@@ -0,0 +1,113 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "अंतिम आवाज क्लोनिंग उपकरण, बेजोड़ शक्ति, प्रतिरूपकता और उपयोगकर्ता के अनुकूल अनुभव के लिए सावधानीपूर्वक अनुकूलित।",
+ "This section contains some extra utilities that often may be in experimental phases.": "इस खंड में कुछ अतिरिक्त उपयोगिताओं हैं जो अक्सर प्रयोगात्मक चरणों में हो सकती हैं।",
+ "Output Information": "आउटपुट जानकारी",
+ "Inference": "अनुमान",
+ "Train": "रेलगाड़ी",
+ "Extra": "अति",
+ "Merge Audios": "ऑडियो मर्ज करें",
+ "Processing": "संसाधन",
+ "Audio Analyzer": "ऑडियो विश्लेषक",
+ "Model Information": "मॉडल जानकारी",
+ "Download": "डाउनलोड",
+ "Report a Bug": "बग की रिपोर्ट करें",
+ "Preprocess": "प्रीप्रोसेस",
+ "Model Name": "मॉडल का नाम",
+ "Enter model name": "मॉडल का नाम दर्ज करें",
+ "Dataset Path": "डेटासेट पथ",
+ "Enter dataset path": "डेटासेट पथ दर्ज करें",
+ "Sampling Rate": "नमूनाकरण दर",
+ "RVC Version": "RVC संस्करण",
+ "Preprocess Dataset": "प्रीप्रोसेस डेटासेट",
+ "Extract": "निकालना",
+ "Hop Length": "हॉप लंबाई",
+ "Batch Size": "बैच का आकार",
+ "Save Every Epoch": "हर युग को बचाओ",
+ "Total Epoch": "कुल युग",
+ "Pretrained": "पूर्व-प्रशिक्षित",
+ "Save Only Latest": "केवल नवीनतम सहेजें",
+ "Save Every Weights": "हर वजन बचाओ",
+ "Custom Pretrained": "कस्टम पूर्व-प्रशिक्षित",
+ "Upload Pretrained Model": "पूर्व-प्रशिक्षित मॉडल अपलोड करें",
+ "Pretrained Custom Settings": "पूर्व-प्रशिक्षित कस्टम सेटिंग्स",
+ "The file you dropped is not a valid pretrained file. Please try again.": "आपके द्वारा छोड़ी गई फ़ाइल कोई मान्य पूर्व-प्रशिक्षित फ़ाइल नहीं है. कृपया पुनः प्रयास करें.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "ड्रॉपडाउन मेनू में पूर्व-प्रशिक्षित फ़ाइल देखने के लिए ताज़ा करें बटन पर क्लिक करें।",
+ "Pretrained G Path": "कस्टम पूर्व-प्रशिक्षित जी",
+ "Pretrained D Path": "कस्टम पूर्वप्रशिक्षित डी",
+ "GPU Settings": "GPU सेटिंग्स",
+ "GPU Custom Settings": "GPU कस्टम सेटिंग्स",
+ "GPU Number": "GPU नंबर",
+ "0 to ∞ separated by -": "0 से ∞ द्वारा अलग किया गया -",
+ "GPU Information": "GPU सूचना",
+ "Pitch Guidance": "पिच मार्गदर्शन",
+ "Extract Features": "एक्सट्रैक्ट फीचर्स",
+ "Start Training": "प्रशिक्षण शुरू करें",
+ "Generate Index": "इंडेक्स जनरेट करें",
+ "Voice Model": "आवाज मॉडल",
+ "Index File": "अनुक्रमणिका फ़ाइल",
+ "Refresh": "आराम देना",
+ "Unload Voice": "आवाज उतारना",
+ "Single": "अकेला",
+ "Upload Audio": "ऑडियो अपलोड करें",
+ "Select Audio": "ऑडियो का चयन करें",
+ "Advanced Settings": "उन्नत सेटिंग्स",
+ "Clear Outputs (Deletes all audios in assets/audios)": "आउटपुट साफ़ करें (संपत्ति/ऑडियो में सभी ऑडियो हटाता है)",
+ "Custom Output Path": "कस्टम आउटपुट पथ",
+ "Output Path": "आउटपुट पथ",
+ "Pitch": "फेंकना",
+ "If >=3: apply median filtering to the harvested pitch results. The value represents the filter radius and can reduce breathiness": "यदि > = 3: कटाई पिच परिणामों के लिए माध्यिका फ़िल्टरिंग लागू करें। मान फिल्टर त्रिज्या का प्रतिनिधित्व करता है और सांस लेने को कम कर सकता है",
+ "Search Feature Ratio": "खोज सुविधा अनुपात",
+ "Pitch extraction algorithm": "पिच निष्कर्षण एल्गोरिथ्म",
+ "Convert": "बदलना",
+ "Export Audio": "ऑडियो निर्यात करें",
+ "Batch": "जत्था",
+ "Input Folder": "इनपुट फ़ोल्डर",
+ "Enter input path": "इनपुट पथ दर्ज करें",
+ "Output Folder": "आउटपुट फ़ोल्डर",
+ "Enter output path": "आउटपुट पथ दर्ज करें",
+ "Get information about the audio": "ऑडियो के बारे में जानकारी प्राप्त करें",
+ "Information about the audio file": "ऑडियो फ़ाइल के बारे में जानकारी",
+ "Waiting for information...": "जानकारी का इंतजार...",
+ "Model fusion": "मॉडल फ्यूजन",
+ "Weight for Model A": "मॉडल ए के लिए वजन",
+ "Whether the model has pitch guidance": "क्या मॉडल में पिच मार्गदर्शन है",
+ "Model architecture version": "मॉडल वास्तुकला संस्करण",
+ "Path to Model A": "मॉडल ए के लिए पथ",
+ "Path to Model B": "मॉडल बी का रास्ता",
+ "Path to model": "मॉडल का मार्ग",
+ "Model information to be placed": "मॉडल जानकारी रखी जाएगी",
+ "Fusion": "परमाणु-संलयन",
+ "Modify model information": "मॉडल जानकारी संशोधित करें",
+ "Path to Model": "मॉडल का मार्ग",
+ "Model information to be modified": "संशोधित की जाने वाली मॉडल जानकारी",
+ "Save file name": "फ़ाइल नाम सहेजें",
+ "Modify": "सुधारना",
+ "View model information": "मॉडल की जानकारी देखें",
+ "View": "देखना",
+ "Model extraction": "मॉडल निष्कर्षण",
+ "Model conversion": "मॉडल रूपांतरण",
+ "Pth file": "Pth फ़ाइल",
+ "Output of the pth file": "pth फ़ाइल का आउटपुट",
+ "# How to Report an Issue on GitHub": "# GitHub पर किसी समस्या की रिपोर्ट कैसे करें",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. आप जिस समस्या का अनुभव कर रहे हैं उसे रिकॉर्ड करना शुरू करने के लिए नीचे दिए गए 'रिकॉर्ड स्क्रीन' बटन पर क्लिक करें।",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. एक बार जब आप समस्या को रिकॉर्ड करना समाप्त कर लेते हैं, तो 'स्टॉप रिकॉर्डिंग' बटन पर क्लिक करें (वही बटन, लेकिन लेबल इस बात पर निर्भर करता है कि आप सक्रिय रूप से रिकॉर्डिंग कर रहे हैं या नहीं)।",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub Issues](https://github.com/IAHispano/Applio/issues) पर जाएं और 'नया मुद्दा' बटन पर क्लिक करें।",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. प्रदान किए गए मुद्दे टेम्पलेट को पूरा करें, आवश्यकतानुसार विवरण शामिल करना सुनिश्चित करें, और पिछले चरण से रिकॉर्ड की गई फ़ाइल को अपलोड करने के लिए संपत्ति अनुभाग का उपयोग करें।",
+ "Record Screen": "रिकॉर्ड स्क्रीन",
+ "Record": "रिकॉर्ड",
+ "Stop Recording": "रिकॉर्डिंग बंद करो",
+ "Introduce the model .pth path": "मॉडल .pth पथ का परिचय दें",
+ "See Model Information": "मॉडल जानकारी देखें",
+ "## Download Model": "## मॉडल डाउनलोड करें",
+ "Model Link": "मॉडल लिंक",
+ "Introduce the model link": "मॉडल लिंक का परिचय दें",
+ "Download Model": "Descargar Modelo",
+ "## Drop files": "## फ़ाइलें ड्रॉप करें",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "अपनी .pth फ़ाइल और .index फ़ाइल को इस स्थान में खींचें. एक को खींचें और फिर दूसरे को।",
+ "TTS Voices": "टीटीएस आवाज़ें",
+ "Text to Synthesize": "संश्लेषित करने के लिए पाठ",
+ "Enter text to synthesize": "संश्लेषित करने के लिए पाठ दर्ज करें",
+ "Output Path for TTS Audio": "TTS ऑडियो के लिए आउटपुट पथ",
+ "Output Path for RVC Audio": "RVC ऑडियो के लिए आउटपुट पथ",
+ "Enable Applio integration with Discord presence": "Applio प्रसेन्स"
+}
diff --git a/assets/i18n/languages/hi_IN.json b/assets/i18n/languages/hi_IN.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc438f38b1f0fb9ded02cffaecbe1877ab1833ef
--- /dev/null
+++ b/assets/i18n/languages/hi_IN.json
@@ -0,0 +1,215 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "अल्टीमेट वॉयस क्लोनिंग टूल, बेजोड़ पॉवर, मॉड्यूलरिटी और उपयोगकर्ता-अनुकूल अनुभव के लिए बारीकी से ऑप्टिमाइज़ किया गया।\n[हिन्दी अनुवाद: Enes](https://discord.com/users/1140031358006202468)",
+ "This section contains some extra utilities that often may be in experimental phases.": "इस खंड में कुछ अतिरिक्त उपयोगिताएँ होती हैं जो अक्सर प्रायोगिक चरणों में हो सकती हैं।",
+ "Output Information": "आउटपुट जानकारी",
+ "The output information will be displayed here.": "आउटपुट जानकारी यहाँ प्रदर्शित की जाएगी।",
+
+ "Inference": "निष्कर्ष",
+ "Train": "ट्रेन",
+ "Extra": "अतिरिक्त",
+ "Merge Audios": "इस ऑडियो को मर्ज करें",
+ "Processing": "प्रोसेसिंग",
+ "Audio Analyzer": "ऑडियो एनालाइज़र",
+ "Model Information": "मॉडल जानकारी",
+ "Plugins": "प्लगइन्स",
+ "Download": "डाउनलोड करें",
+ "Report a Bug": "एक बग की रिपोर्ट करें",
+ "Settings": "सेटिंग्स",
+
+ "Preprocess": "पूर्व-प्रसंस्करण",
+ "Model Name": "मॉडल का नाम",
+ "Name of the new model.": "नए मॉडल का नाम।",
+ "Enter model name": "मॉडल नाम डालें",
+ "Dataset Path": "डेटासेट पथ",
+ "Path to the dataset folder.": "डेटासेट फ़ोल्डर का पथ।",
+ "Refresh Datasets": "डेटासेट रीफ्रेश करें",
+ "Dataset Creator": "डेटासेट बनाने वाला",
+ "Dataset Name": "डेटासेट का नाम",
+ "Name of the new dataset.": "नए डेटासेट का नाम।",
+ "Enter dataset name": "डेटासेट का नाम डालें",
+ "Upload Audio Dataset": "ऑडियो डेटासेट अपलोड करें",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ऑडियो फ़ाइल को डेटासेट में सफलतापूर्वक जोड़ा गया है। कृपया प्रीप्रोसेस बटन पर क्लिक करें।",
+ "Enter dataset path": "डेटासेट पथ डालें",
+ "Sampling Rate": "नमूनाकरण दर",
+ "The sampling rate of the audio files.": "ऑडियो फ़ाइलों की नमूनाकरण दर।",
+ "RVC Version": "RVC वर्शन",
+ "The RVC version of the model.": "मॉडल का RVC वर्शन।",
+ "Preprocess Dataset": "डेटासेट का पूर्व-प्रसंस्करण करें",
+
+ "Embedder Model": "एम्बेडर मॉडल",
+ "Model used for learning speaker embedding.": "स्पीकर एम्बेडिंग सीखने के लिए उपयोग किया जाने वाला मॉडल।",
+ "Extract": "एक्सट्रैक्ट",
+ "Hop Length": "हॉप लंबाई",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "यह अवधी को दर्शाती है जिसे सिस्टम को पिच में महत्वपूर्ण बदलाव के लिए ले जाना पड़ता है। कम हॉप लंबाई को अनुमान लगाने में अधिक समय लगता है लेकिन उच्च पिच सटीकता मिलती है।",
+ "Batch Size": "बैच आकार",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "अपने GPU की उपलब्ध VRAM के साथ इसे संरेखित करना उचित है। 4 की सेटिंग में सटीकता बेहतर होती है लेकिन प्रोसेसिंग धीमी होती है, जबकि 8 तेज और मानक परिणाम प्रदान करता है।",
+ "Save Every Epoch": "प्रत्येक युग को सेव करें",
+ "Determine at how many epochs the model will saved at.": "निर्धारित करें कि कितने युग पर मॉडल सहेजा जाएगा",
+ "Total Epoch": "कुल युग",
+ "Specifies the overall quantity of epochs for the model training process.": "मॉडल प्रशिक्षण प्रक्रिया के लिए युग की समग्र मात्रा निर्दिष्ट करता है।",
+ "Pretrained": "पूर्व प्रशिक्षित",
+ "Save Only Latest": "केवल नवीनतम को सेव करें",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "इस सेटिंग को सक्षम करने पर G और D फ़ाइलें अपने केवल नवीनतम संस्करण को ही सेव करेंगी, भंडारण स्थान को प्रभावी ढंग से संरक्षित करना।",
+ "Save Every Weights": "प्रत्येक वज़न को सेव करें",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "यह सेटिंग आपको प्रत्येक युग के समापन पर मॉडल के वजन को सेव करने में सक्षम बनाती है।",
+ "Custom Pretrained": "कस्टम पूर्व प्रशिक्षित",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "कस्टम पूर्व प्रशिक्षित मॉडलों का उपयोग करने से बेहतर परिणाम मिल सकते हैं, क्योंकि विशिष्ट उपयोग के मामले के अनुरूप सबसे उपयुक्त पूर्व प्रशिक्षित मॉडलों को चुनने से प्रदर्शन में काफी वृद्धि हो सकती है।",
+ "Upload Pretrained Model": "पूर्व प्रशिक्षित मॉडल अपलोड करें",
+ "Refresh Custom Pretraineds": "कस्टम पूर्व प्रशिक्षितों को ताज़ा करें",
+ "Pretrained Custom Settings": "पूर्व प्रशिक्षित कस्टम सेटिंग्स",
+ "The file you dropped is not a valid pretrained file. Please try again.": "जो फ़ाइल आपने छोड़ी है वह एक मान्य पूर्व प्रशिक्षित फ़ाइल नहीं है। कृपया पुनः प्रयास करें।",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "ड्रॉपडाउन मेनू में पूर्व प्रशिक्षित फ़ाइल को देखने के लिए रीफ़्रेश बटन पर क्लिक करें।",
+ "Pretrained G Path": "कस्टम पूर्व प्रशिक्षित G",
+ "Pretrained D Path": "कस्टम पूर्व प्रशिक्षित D",
+ "GPU Settings": "GPU सेटिंग्स",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "उन्नत GPU सेटिंग्स सेट करता है, बेहतर GPU आर्किटेक्चर वाले उपयोगकर्ताओं के लिए अनुशंसित।",
+ "GPU Custom Settings": "GPU कस्टम सेटिंग्स",
+ "GPU Number": "GPU संख्या",
+ "0 to ∞ separated by -": "0 से ∞ तक अलग से -",
+ "GPU Information": "GPU जानकारी",
+ "Pitch Guidance": "पिच मार्गदर्शन",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "पिच मार्गदर्शन को नियोजित करके, मूल आवाज़ के स्वर को प्रतिबिंबित करना संभव हो जाता है। यह सुविधा गायन और अन्य परिदृश्य के लिए विशेष रूप से मूल्यवान है जहाँ मूल राग या पिच पैटर्न को संरक्षित करना आवश्यक है।",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "अपना खुद का प्रशिक्षण करते समय पूर्व प्रशिक्षित मॉडलों का उपयोग करें। यह दृष्टिकोण प्रशिक्षण की अवधि कम करता है और समग्र गुणवत्ता को बढ़ाता है।",
+ "Extract Features": "विशेषताएँ निकालें",
+ "Overtraining Detector": "ओवरट्रेनिंग डिटेक्टर",
+ "Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data.": "मॉडल को प्रशिक्षण डेटा को बहुत अच्छी तरह से सीखने और नए डेटा को सामान्य बनाने की क्षमता खोने से रोकने के लिए ओवरट्रेनिंग का पता लगाएं।",
+ "Overtraining Detector Settings": "ओवरट्रेनिंग डिटेक्टर सेटिंग्स",
+ "Overtraining Threshold": "ओवरट्रेनिंग थ्रेशोल्ड",
+ "Set the maximum number of epochs you want your model to stop training if no improvement is detected.": "यदि कोई सुधार नहीं पाया जाता है, तो आप अपने मॉडल को प्रशिक्षण बंद करने के लिए अधिकतम युगों की संख्या निर्धारित करें।",
+
+ "Start Training": "प्रशिक्षण शुरू करें",
+ "Stop Training & Restart Applio": "प्रशिक्षण रोकें और Applio को पुनः आरंभ करें",
+ "Generate Index": "इंडेक्स बनाएँ",
+
+ "Export Model": "मॉडल निर्यात करें",
+ "The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive.": "'अपलोड' बटन केवल गूगल कोलाब के लिए है: निर्यात की गई फ़ाइलों को आपके Google ड्राइव में ApplioExported फ़ोल्डर में अपलोड करता है।",
+ "Exported Pth file": "निर्यात की गई Pth फ़ाइल",
+ "Exported Index file": "निर्यात की गई इंडेक्स फ़ाइल",
+ "Select the pth file to be exported": "निर्यात की जाने वाली pth फ़ाइल का चयन करें",
+ "Select the index file to be exported": "निर्यात की जाने वाली इंडेक्स फ़ाइल का चयन करें",
+ "Upload": "अपलोड करें",
+
+ "Voice Model": "आवाज़ का मॉडल",
+ "Select the voice model to use for the conversion.": "रूपांतरण के लिए उपयोग करने के लिए आवाज मॉडल का चयन करें।",
+ "Index File": "इंडेक्स फ़ाइल",
+ "Select the index file to use for the conversion.": "रूपांतरण के लिए उपयोग करने के लिए इंडेक्स फ़ाइल का चयन करें।",
+ "Refresh": "ताज़ा करें",
+ "Unload Voice": "आवाज़ अनलोड करें",
+
+ "Single": "सिंगल",
+ "Upload Audio": "ऑडियो अपलोड करें",
+ "Select Audio": "ऑडियो चुनें",
+ "Select the audio to convert.": "रूपांतरित करने के लिए ऑडियो चुनें।",
+ "Advanced Settings": "उन्नत सेटिंग्स",
+ "Clear Outputs (Deletes all audios in assets/audios)": "(assets/audios में सभी ऑडियो को हटाता है) आउटपुट साफ़ करें",
+ "Custom Output Path": "कस्टम आउटपुट पथ",
+ "Output Path": "आउटपुट पथ",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "आउटपुट ऑडियो का पथ जहाँ उसे सेव किया जाएगा, डिफ़ॉल्ट रूप से assets/audios/output.wav में",
+ "Split Audio": "ऑडियो को विभाजित करें",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "कुछ स्थितियों में अधिक बेहतर परिणाम प्राप्त करने के लिए अनुमान के लिए ऑडियो को हिस्सों में विभाजित करें।",
+ "Autotune": "ऑटोट्यून",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "अपने अनुमानों पर एक नरम ऑटोट्यून लागू करें, यह गायन रूपांतरण के लिए अनुशंसित है।",
+ "Clean Audio": "ऑडियो साफ़ करें",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "शोर का पता लगाने वाले एल्गोरिदम का उपयोग करके अपने ऑडियो आउटपुट को साफ़ करें, बोलने वाले ऑडियो के लिए अनुशंसित।",
+ "Clean Strength": "सफाई शक्ति",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "ऑडियो को साफ़ करने का स्तर निर्धारित करें, जैसे-जैसे आप इसे बढ़ाते जाएँगे यह उतना ही अधिक साफ़ करेगा, लेकिन यह संभव है कि ऑडियो और संकुचित हो सकता है।",
+ "Pitch": "पिच",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ऑडियो की पिच सेट करें, मान जितना अधिक होगा, पिच उतनी ही अधिक होगी।",
+ "Filter Radius": "फ़िल्टर त्रिज्या",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "यदि संख्या तीन से अधिक या उसके बराबर है, तो एकत्र किए गए स्वर परिणामों पर मेडियन फ़िल्टरिंग का उपयोग करके साँस लेना कम हो जाता है।",
+ "Search Feature Ratio": "फीचर अनुपात खोजें",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "इंडेक्स फ़ाइल द्वारा प्रभावित; एक उच्च मान अधिक प्रभाव से मेल खाता है। हालाँकि, कम मूल्यों को चुनने से ऑडियो में मौजूद कलाकृतियों को कम करने में मदद मिल सकती है।",
+ "Volume Envelope": "आयतन आवरण",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "आउटपुट के आयतन आवरण के साथ स्थानापन्न करें या सम्मिश्रित करें। अनुपात जितना 1 के करीब होगा, आउटपुट आवरण उतना ही अधिक नियोजित किया जाएगा।",
+ "Protect Voiceless Consonants": "वॉयसलेस व्यंजनों की सुरक्षा करें",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "इलेक्ट्रो-ध्वनिक फाड़ और अन्य कलाकृतियों को रोकने के लिए अलग-अलग व्यंजन और साँस लेने की आवाज़ को सुरक्षित रखें। पैरामीटर को उसके अधिकतम मान 0.5 तक खींचना व्यापक सुरक्षा प्रदान करता है। हालाँकि, इस मान को कम करने से सुरक्षा की सीमा कम हो सकती है जबकि संभावित रूप से अनुक्रमण प्रभाव कम हो सकता है।",
+ "Pitch extraction algorithm": "पिच निष्कर्षण एल्गोरिदम",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ऑडियो रूपांतरण के लिए उपयोग करने के लिए पिच निष्कर्षण एल्गोरिदम। डिफ़ॉल्ट एल्गोरिथम rmvpe है, जो अधिकांश मामलों के लिए अनुशंसित है।",
+
+ "Convert": "रूपांतरित करें",
+ "Export Audio": "ऑडियो एक्सपोर्ट करें",
+
+ "Batch": "बैच",
+ "Input Folder": "इनपुट फ़ोल्डर",
+ "Select the folder containing the audios to convert.": "रूपांतरित करने के लिए ऑडियो वाली फ़ोल्डर का चयन करें।",
+ "Enter input path": "इनपुट पथ दर्ज करें",
+ "Output Folder": "आउटपुट फ़ोल्डर",
+ "Select the folder where the output audios will be saved.": "फ़ोल्डर का चयन करें जहाँ आउटपुट ऑडियो को सेव किया जाएगा।",
+ "Enter output path": "आउटपुट पथ दर्ज करें",
+
+ "Get information about the audio": "ऑडियो के बारे में जानकारी प्राप्त करें",
+
+ "## Voice Blender": "## Voice Blender",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "दो आवाज़ के मॉडल का चयन करें, अपना वांछित मिश्रित प्रतिशत सेट करें, और उन्हें एक पूर्ण रूप से नई आवाज़ में सम्मिश्रित करें।",
+ "Voice Blender": "Voice Blender",
+ "Drag and drop your model here": "यहाँ अपना मॉडल ड्रैग एंड ड्रॉप करें",
+ "You can also use a custom path.": "आप कस्टम पथ का भी उपयोग कर सकते हैं।",
+ "Blend Ratio": "मिश्रित अनुपात",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "स्थिति को एक तरफ या दूसरी तरफ समायोजित करने से माॅडल पहले या दूसरे के समान हो जाएगा।",
+ "Fusion": "सम्मिश्रण",
+
+ "Path to Model": "मॉडल का पथ",
+ "Enter path to model": "मॉडल का पथ दर्ज करें",
+ "Model information to be placed": "मॉडल की जानकारी स्थान की जाएगी",
+ "Inroduce the model information": "मॉडल की जानकारी का परिचय दें।",
+ "The information to be placed in the model (You can leave it blank or put anything).": "मॉडल में रखी जाने वाली जानकारी (आप इसे खाली छोड़ सकते हैं या कुछ भी डाल सकते हैं)।",
+ "View model information": "मॉडल की जानकारी देखें",
+ "Introduce the model pth path": "मॉडल pth पथ का परिचय दें",
+ "View": "देखें",
+ "Model extraction": "मॉडल निष्कर्षण",
+ "Model conversion": "मॉडल रूपांतरण",
+ "Pth file": "Pth फ़ाइल",
+ "Output of the pth file": "Pth फ़ाइल का आउटपुट",
+
+ "# How to Report an Issue on GitHub": "# GitHub पर किसी समस्या की रिपोर्ट कैसे करें",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1.आपके द्वारा अनुभव की जा रही समस्या को रिकॉर्ड करना शुरू करने के लिए नीचे दिए गए 'रिकॉर्ड स्क्रीन' बटन पर क्लिक करें।",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. एक बार जब आप समस्या की रिकॉर्डिंग समाप्त कर लें, तो 'स्टॉप रिकॉर्डिंग' बटन पर क्लिक करें (वही बटन, लेकिन लेबल इस पर निर्भर करता है कि आप सक्रिय रूप से रिकॉर्डिंग कर रहे हैं या नहीं)।",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub Issues](https://github.com/IAHispano/Applio/issues) पर जाएँ और 'New Issue' बटन पर क्लिक करें।",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. दिए गए समस्या टेम्पलेट को पूरा करें, आवश्यकतानुसार विवरण शामिल करना सुनिश्चित करें, और पिछले चरण से रिकॉर्ड की गई फ़ाइल को अपलोड करने के लिए एसेट सेक्शन का उपयोग करें।",
+
+ "Record Screen": "स्क्रीन रिकॉर्ड करें",
+ "Record": "रिकॉर्ड करें",
+ "Stop Recording": "रिकॉर्डिंग बंद करें",
+
+ "Introduce the model .pth path": "मॉडल .pth पथ का परिचय दें",
+ "See Model Information": "मॉडल की जानकारी देखें",
+
+ "## Download Model": "## मॉडल डाउनलोड करें",
+ "Model Link": "मॉडल लिंक",
+ "Introduce the model link": "मॉडल लिंक का परिचय दें",
+ "Download Model": "मॉडल डाउनलोड करें",
+ "## Drop files": "## ड्रॉप फ़ाइलें",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "अपनी .pth फ़ाइल और .index फ़ाइल को इस स्थान पर ड्रैग करें। एक को ड्रैग करें और फिर दूसरे को ड्रैग करें।",
+ "## Search Model": "## मॉडल खोजें",
+ "Search": "खोज",
+ "Introduce the model name to search.": "खोज करने के लिए मॉडल का नाम बताएं।",
+ "We couldn't find models by that name.": "हमें उस नाम से मॉडल नहीं मिला।",
+
+ "TTS Voices": "TTS Voices",
+ "Select the TTS voice to use for the conversion.": "रूपांतरण के लिए उपयोग करने के लिए TTS वॉयस का चयन करें।",
+ "Text to Synthesize": "Text to Synthesize",
+ "Enter the text to synthesize.": "संश्लेषित करने के लिए पाठ दर्ज करें।",
+ "Or you can upload a .txt file": "या आप एक .txt फ़ाइल अपलोड कर सकते हैं",
+ "Enter text to synthesize": "संश्लेषित करने के लिए पाठ दर्ज करें",
+ "Output Path for TTS Audio": "TTS ऑडियो के लिए आउटपुट पथ",
+ "Output Path for RVC Audio": "RVC ऑडियो के लिए आउटपुट पथ",
+
+ "Enable Applio integration with Discord presence": "डिस्कॉर्ड उपस्थिति के साथ Applio एकीकरण को सक्षम करें",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "यह डिस्कॉर्ड में वर्तमान Applio गतिविधि प्रदर्शित करने की संभावना को सक्रिय करेगा।",
+ "Enable Applio integration with applio.org/models using flask": "flask का उपयोग करके applio.org/models के साथ Applio एकीकरण को सक्षम करें",
+ "It will activate the possibility of downloading models with a click from the website.": "यह वेबसाइट से एक क्लिक से मॉडल डाउनलोड करने की संभावना को सक्रिय करेगा।",
+ "Enable fake GPU": "नकली GPU सक्षम करें",
+ "Training is currently unsupported due to the absence of a GPU. To activate the training tab, navigate to the settings tab and enable the 'Fake GPU' option.": "GPU की अनुपस्थिति के कारण प्रशिक्षण वर्तमान में असमर्थित है। प्रशिक्षण टैब को सक्रिय करने के लिए, सेटिंग टैब पर नेविगेट करें और 'नकली GPU' विकल्प को सक्षम करें।",
+ "Activates the train tab. However, please note that this device lacks GPU capabilities, hence training is not supported. This option is only for testing purposes. (This option will restart Applio)": "ट्रेन टैब को सक्रिय करता है। हालांकि, कृपया ध्यान दें कि इस डिवाइस में GPU क्षमताएं नहीं हैं, इसलिए प्रशिक्षण समर्थित नहीं है। यह विकल्प केवल परीक्षण उद्देश्यों के लिए है। (यह विकल्प Applio को पुनरारंभ करेगा)",
+ "Theme": "थीम",
+ "Select the theme you want to use. (Requires restarting Applio)": "वह थीम चुनें जिसका आप उपयोग करना चाहते हैं। (Applio को पुनरारंभ करने की आवश्यकता है)",
+ "Language": "भाषा",
+ "Select the language you want to use. (Requires restarting Applio)": "वह भाषा चुनें जिसका आप उपयोग करना चाहते हैं। (Applio को पुनरारंभ करने की आवश्यकता है)",
+
+ "Plugin Installer": "प्लगइन इंस्टॉलर",
+ "Drag your plugin.zip to install it": "इसे इंस्टॉल करने के लिए अपने plugin.zip को ड्रैग करें",
+
+ "Version Checker": "संस्करण चेकर",
+ "Check which version of Applio is the latest to see if you need to update.": "अद्यतित करने के लिए आपको चाहिए या नहीं यह देखने के लिए जाँच करें कि Applio का कौन सा संस्करण नवीनतम है।",
+ "Check for updates": "अपडेट जांचें"
+}
diff --git a/assets/i18n/languages/hu_HU.json b/assets/i18n/languages/hu_HU.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8b7e5bac2986b8d43f83e1523bd27b4ef2410b5
--- /dev/null
+++ b/assets/i18n/languages/hu_HU.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "A legjobb hangklónozó eszköz, aprólékosan optimalizálva a páratlan teljesítmény, a modularitás és a felhasználóbarát élmény érdekében.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Ez a szakasz néhány további segédprogramot tartalmaz, amelyek gyakran kísérleti fázisban vannak.",
+ "Output Information": "Kimeneti információk",
+ "The output information will be displayed here.": "A kimeneti információk itt jelennek meg.",
+ "Inference": "Következtetés",
+ "Train": "Vonat",
+ "Extra": "Többlet",
+ "Merge Audios": "Hangok egyesítése",
+ "Processing": "Feldolgozás",
+ "Audio Analyzer": "Hangelemző",
+ "Model Information": "Modell információk",
+ "Plugins": "Bővítmények",
+ "Download": "Letöltés",
+ "Report a Bug": "Hiba jelentése",
+ "Settings": "Beállítások",
+ "Preprocess": "Előfeldolgozás",
+ "Model Name": "Modell neve",
+ "Name of the new model.": "Az új modell neve.",
+ "Enter model name": "Adja meg a modell nevét",
+ "Dataset Path": "Adatkészlet elérési útja",
+ "Path to the dataset folder.": "Az adatkészletmappa elérési útja.",
+ "Refresh Datasets": "Adatkészletek frissítése",
+ "Dataset Creator": "Adatkészlet létrehozója",
+ "Dataset Name": "Adatkészlet neve",
+ "Name of the new dataset.": "Az új adatkészlet neve.",
+ "Enter dataset name": "Adja meg az adatkészlet nevét",
+ "Upload Audio Dataset": "Hangadatkészlet feltöltése",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "A hangfájl sikeresen hozzá lett adva az adatkészlethez. Kérem kattintson az előfeldolgozás gombra.",
+ "Enter dataset path": "Adja meg az adatkészlet elérési útját",
+ "Sampling Rate": "Mintavételi arány",
+ "The sampling rate of the audio files.": "Az audiofájlok mintavételi frekvenciája.",
+ "RVC Version": "RVC verzió",
+ "The RVC version of the model.": "A modell RVC verziója.",
+ "Preprocess Dataset": "Adatkészlet előfeldolgozása",
+ "Extract": "Kivonat",
+ "Hop Length": "Komló hossza",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Azt az időtartamot jelöli, amely alatt a rendszer jelentős hangmagasság-változásra vált. A kisebb ugráshosszak több időt igényelnek a következtetéshez, de általában nagyobb hangmagasság-pontosságot eredményeznek.",
+ "Batch Size": "Tétel mérete",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Javasoljuk, hogy igazítsa a GPU rendelkezésre álló VRAM-jához. A 4-es beállítás nagyobb pontosságot, de lassabb feldolgozást biztosít, míg a 8-as gyorsabb és szabványos eredményeket biztosít.",
+ "Save Every Epoch": "Mentsd meg minden korszakot",
+ "Determine at how many epochs the model will saved at.": "Határozza meg, hogy hány korszakban menti a modellt.",
+ "Total Epoch": "Teljes korszak",
+ "Specifies the overall quantity of epochs for the model training process.": "A modell betanítási folyamatának epocháinak teljes mennyiségét adja meg.",
+ "Pretrained": "Előre betanított",
+ "Save Only Latest": "Csak a legújabbak mentése",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "A beállítás engedélyezése azt eredményezi, hogy a G és D fájlok csak a legújabb verziójukat mentik, így hatékonyan megtakarítják a tárhelyet.",
+ "Save Every Weights": "Takarítson meg minden súlyt",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Ez a beállítás lehetővé teszi a modell súlyozásának mentését az egyes korszakok végén.",
+ "Custom Pretrained": "Egyéni előképzett",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Az egyéni előre betanított modellek használata kiváló eredményekhez vezethet, mivel az adott használati esetre szabott legmegfelelőbb előre betanított modellek kiválasztása jelentősen javíthatja a teljesítményt.",
+ "Upload Pretrained Model": "Előre betanított modell feltöltése",
+ "Refresh Custom Pretraineds": "Egyéni előképzetek frissítése",
+ "Pretrained Custom Settings": "Előre betanított egyéni beállítások",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Az eldobott fájl nem érvényes előre betanított fájl. Kérjük, próbálja újra.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Kattintson a frissítés gombra az előre betanított fájl megjelenítéséhez a legördülő menüben.",
+ "Pretrained G Path": "Egyéni előképzett G",
+ "Pretrained D Path": "Egyéni előképzett D",
+ "GPU Settings": "GPU-beállítások",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Speciális GPU-beállításokat állít be, amelyek a jobb GPU-architektúrával rendelkező felhasználók számára ajánlottak.",
+ "GPU Custom Settings": "GPU egyéni beállítások",
+ "GPU Number": "GPU-szám",
+ "0 to ∞ separated by -": "0-tól ∞-ig - választja el",
+ "GPU Information": "GPU-információk",
+ "Pitch Guidance": "Pitch útmutatás",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "A hangmagasság-útmutatás alkalmazásával megvalósíthatóvá válik az eredeti hang intonációjának tükrözése, beleértve annak hangmagasságát is. Ez a funkció különösen értékes énekléshez és más forgatókönyvekhez, ahol az eredeti dallam vagy hangmagasság minta megőrzése elengedhetetlen.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Használja az előre betanított modelleket a saját betanításakor. Ez a megközelítés csökkenti a képzés időtartamát és javítja az általános minőséget.",
+ "Extract Features": "Jellemzők kivonása",
+ "Start Training": "Kezdje el az edzést",
+ "Generate Index": "Index létrehozása",
+ "Voice Model": "Hangmodell",
+ "Select the voice model to use for the conversion.": "Válassza ki az átalakításhoz használni kívánt hangmodellt.",
+ "Index File": "Index fájl",
+ "Select the index file to use for the conversion.": "Válassza ki az átalakításhoz használni kívánt indexfájlt.",
+ "Refresh": "Felfrissít",
+ "Unload Voice": "Hang eltávolítása",
+ "Single": "Nőtlen",
+ "Upload Audio": "Hang feltöltése",
+ "Select Audio": "Válassza az Audio lehetőséget",
+ "Select the audio to convert.": "Válassza ki a konvertálni kívánt hangot.",
+ "Advanced Settings": "Speciális beállítások",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Kimenetek törlése (Törli az összes hangot az eszközökből/hangokból)",
+ "Custom Output Path": "Egyéni kimeneti útvonal",
+ "Output Path": "Kimeneti útvonal",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Az elérési út, ahová a kimeneti hang mentésre kerül, alapértelmezés szerint az eszközökben / hangokban / output.wav",
+ "Split Audio": "Osztott hang",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Ossza fel a hangot darabokra a következtetéshez, hogy bizonyos esetekben jobb eredményeket érjen el.",
+ "Autotune": "Automatikus hangolás",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Alkalmazzon lágy automatikus hangolást a következtetésekre, ami az énekkonverziókhoz ajánlott.",
+ "Clean Audio": "Tiszta hangzás",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Tisztítsa meg a hangkimenetet zajérzékelő algoritmusokkal, amelyek a hangok beszédéhez ajánlottak.",
+ "Clean Strength": "Tiszta szilárdság",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Állítsa be a tisztítási szintet a kívánt hangra, minél jobban növeli, annál jobban megtisztítja, de lehetséges, hogy a hang tömörítettebb lesz.",
+ "Pitch": "Hangmagasság",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Állítsa be a hang hangmagasságát, minél magasabb az érték, annál magasabb a hangmagasság.",
+ "Filter Radius": "Szűrési sugár",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Ha a szám nagyobb vagy egyenlő hárommal, az összegyűjtött tónuseredmények medián szűrésének alkalmazása csökkentheti a légzést.",
+ "Search Feature Ratio": "Keresési funkciók aránya",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Az indexfájl által gyakorolt hatás; A magasabb érték nagyobb befolyásnak felel meg. Az alacsonyabb értékek választása azonban segíthet enyhíteni a hangban jelen lévő műtermékeket.",
+ "Volume Envelope": "Térfogat boríték",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Helyettesítse vagy keverje össze a kimenet térfogatburkológörbéjét. Minél közelebb van az arány az 1-hez, annál nagyobb a kimeneti burkológörbe.",
+ "Protect Voiceless Consonants": "Zöngétlen mássalhangzók védelme",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Védje a különálló mássalhangzókat és légzési hangokat, hogy megakadályozza az elektroakusztikus szakadást és más műtermékeket. A paraméter maximális 0,5-ös értékre való lekérése átfogó védelmet nyújt. Ennek az értéknek a csökkentése azonban csökkentheti a védelem mértékét, miközben potenciálisan enyhítheti az indexelési hatást.",
+ "Pitch extraction algorithm": "Pitch extrakciós algoritmus",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Hangmagasság-kinyerési algoritmus az audio konvertáláshoz. Az alapértelmezett algoritmus az rmvpe, amely a legtöbb esetben ajánlott.",
+ "Convert": "Megtérít",
+ "Export Audio": "Hang exportálása",
+ "Batch": "Halom",
+ "Input Folder": "Bemeneti mappa",
+ "Select the folder containing the audios to convert.": "Válassza ki a konvertálni kívánt hangokat tartalmazó mappát.",
+ "Enter input path": "Adja meg a bemeneti útvonalat",
+ "Output Folder": "Kimeneti mappa",
+ "Select the folder where the output audios will be saved.": "Válassza ki azt a mappát, ahová a kimeneti hangokat menteni kívánja.",
+ "Enter output path": "Adja meg a kimeneti útvonalat",
+ "Get information about the audio": "Információk lekérése a hangról",
+ "Information about the audio file": "Információ a hangfájlról",
+ "Waiting for information...": "Információra várva...",
+ "## Voice Blender": "## Hangos turmixgép",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Válasszon ki két hangmodellt, állítsa be a kívánt keverési százalékot, és keverje össze őket egy teljesen új hangszínbe.",
+ "Voice Blender": "Hangos turmixgép",
+ "Drag and drop your model here": "Húzza ide a modellt",
+ "You can also use a custom path.": "Egyéni elérési utat is használhat.",
+ "Blend Ratio": "Keverési arány",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Ha a pozíciót jobban az egyik vagy a másik oldalra állítja, a modell jobban hasonlít az elsőhöz vagy a másodikhoz.",
+ "Fusion": "Fúzió",
+ "Path to Model": "A modell elérési útja",
+ "Enter path to model": "Adja meg a modell elérési útját",
+ "Model information to be placed": "Az elhelyezendő modellinformációk",
+ "Inroduce the model information": "A modellinformációk bemutatása",
+ "The information to be placed in the model (You can leave it blank or put anything).": "A modellben elhelyezendő információk (üresen hagyhatja, vagy bármit betehet).",
+ "View model information": "Modellinformációk megtekintése",
+ "Introduce the model pth path": "A modell pth elérési útjának bemutatása",
+ "View": "Nézet",
+ "Model extraction": "Modell kinyerése",
+ "Model conversion": "Modell átalakítás",
+ "Pth file": "Pth fájl",
+ "Output of the pth file": "A pth fájl kimenete",
+ "# How to Report an Issue on GitHub": "# Hogyan jelenthet problémát a GitHubon",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Kattintson az alábbi \"Felvétel képernyő\" gombra a tapasztalt probléma rögzítésének megkezdéséhez.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Miután befejezte a probléma rögzítését, kattintson a \"Felvétel leállítása\" gombra (ugyanaz a gomb, de a címke attól függően változik, hogy aktívan rögzít-e vagy sem).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Lépjen a [GitHub-problémák](https://github.com/IAHispano/Applio/issues) oldalra, és kattintson az \"Új probléma\" gombra.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Töltse ki a megadott problémasablont, ügyelve arra, hogy szükség szerint tartalmazza a részleteket, és használja az eszközök szakaszt az előző lépésből rögzített fájl feltöltéséhez.",
+ "Record Screen": "Felvétel képernyő",
+ "Record": "Rekord",
+ "Stop Recording": "Felvétel leállítása",
+ "Introduce the model .pth path": "A modell .pth elérési útjának bemutatása",
+ "See Model Information": "Modellinformációk megtekintése",
+ "## Download Model": "## Modell letöltése",
+ "Model Link": "Modell link",
+ "Introduce the model link": "A modellhivatkozás bemutatása",
+ "Download Model": "Modell letöltése",
+ "## Drop files": "## Dobja el a fájlokat",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Húzza a .pth és .index fájlt erre a helyre. Húzza az egyiket, majd a másikat.",
+ "TTS Voices": "TTS-hangok",
+ "Select the TTS voice to use for the conversion.": "Válassza ki az átalakításhoz használni kívánt TTS-hangot.",
+ "Text to Synthesize": "Szintetizálandó szöveg",
+ "Enter the text to synthesize.": "Írja be a szintetizálni kívánt szöveget.",
+ "Or you can upload a .txt file": "Vagy feltölthet egy .txt fájlt",
+ "Enter text to synthesize": "Írja be a szintetizálni kívánt szöveget",
+ "Output Path for TTS Audio": "A TTS Audio kimeneti útvonala",
+ "Output Path for RVC Audio": "Az RVC Audio kimeneti útvonala",
+ "Enable Applio integration with Discord presence": "Engedélyezze az Applio integrációját a Discord jelenléttel",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Aktiválja az aktuális Applio tevékenység megjelenítésének lehetőségét a Discordban.",
+ "Enable Applio integration with applio.org/models using flask": "Az Applio és a applio.org/models integrációjának engedélyezése lombik használatával",
+ "It will activate the possibility of downloading models with a click from the website.": "Aktiválja a modellek letöltésének lehetőségét egy kattintással a weboldalról.",
+ "Theme": "Téma",
+ "Select the theme you want to use. (Requires restarting Applio)": "Válassza ki a használni kívánt témát. (Az Applio újraindítását igényli)",
+ "Language": "Nyelv",
+ "Select the language you want to use. (Requires restarting Applio)": "Válassza ki a használni kívánt nyelvet. (Az Applio újraindítását igényli)",
+ "Plugin Installer": "Bővítmény telepítő",
+ "Drag your plugin.zip to install it": "Húzza a plugin.zip a telepítéshez",
+ "Version Checker": "Verzióellenőrző",
+ "Check which version of Applio is the latest to see if you need to update.": "Ellenőrizze, hogy az Applio melyik verziója a legújabb, hogy lássa, frissítenie kell-e.",
+ "Check for updates": "Frissítések keresése"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/id_ID.json b/assets/i18n/languages/id_ID.json
new file mode 100644
index 0000000000000000000000000000000000000000..d636ca2cab9b6eea6a7f20d73fc769e5c12961eb
--- /dev/null
+++ b/assets/i18n/languages/id_ID.json
@@ -0,0 +1,148 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Alat kloning suara terbaik, dioptimalkan secara cermat untuk kekuatan tak tertandingi, modularitas, dan pengalaman ramah pengguna.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Bagian ini berisi beberapa utilitas tambahan yang mungkin sering berada dalam tahap percobaan.",
+ "Output Information": "informasi keluaran",
+
+ "Inference": "Inference",
+ "Train": "training model",
+ "Extra": "bonus",
+ "Merge Audios": "Gabungkan Audio",
+ "Processing": "Pengolahan",
+ "Audio Analyzer": "Penganalisis Audio",
+ "Model Information": "informasi",
+ "Plugins": "Plugin",
+ "Download": "Unduh",
+ "Report a Bug": "Laporkan Bug",
+ "Settings": "Pengaturan",
+
+ "Preprocess": "Proses awal",
+ "Model Name": "Nama model",
+ "Enter model name": "masukkan nama model",
+ "Dataset Path": "Jalur Kumpulan Data",
+ "Dataset Creator": "Pembuat Kumpulan Data/dataset",
+ "Dataset Name": "Nama kumpulan data",
+ "Enter dataset name": "Masukkan nama kumpulan data",
+ "Upload Audio Dataset": "Unggah audio kumpulan data",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Berkas audio telah berhasil ditambahkan ke dataset. Silakan klik tombol praproses.",
+ "Enter dataset path": "Masukkan jalur kumpulan data",
+ "Sampling Rate": "Tingkat Pengambilan Sampel",
+ "RVC Version": "Versi RVC",
+ "Preprocess Dataset": "Kumpulan Data Praproses",
+
+ "Extract": "Ekstrak",
+ "Hop Length": "Panjang Lompatan",
+ "Batch Size": "Ukuran Batch",
+ "Save Every Epoch": "Simpan Setiap Epoch",
+ "Total Epoch": "Total Epoch",
+ "Pretrained": "Terlatih sebelumnya",
+ "Save Only Latest": "Simpan Hanya Yang Terbaru",
+ "Save Every Weights": "Simpan Setiap Weights/beban",
+ "Custom Pretrained": "Terlatih Khusus",
+ "Upload Pretrained Model": "Unggah Model yang Telah Dilatih sebelumnya",
+ "Pretrained Custom Settings": "Pengaturan Kustom yang Telah Dilatih Sebelumnya",
+ "The file you dropped is not a valid pretrained file. Please try again.": "File yang Anda jatuhkan bukan file terlatih yang valid. Silakan coba lagi.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Klik tombol segarkan untuk melihat file yang telah dilatih sebelumnya di menu tarik-turun.",
+ "Pretrained G Path": "Terlatih Khusus G",
+ "Pretrained D Path": "Terlatih Khusus D",
+ "GPU Settings": "Penggaturan GPU",
+ "GPU Custom Settings": "Penggaturan GPU khusus",
+ "GPU Number": "Angka GPU",
+ "0 to ∞ separated by -": "0 to ∞ dipisahkan oleh -",
+ "GPU Information": "Informasi GPU",
+ "Pitch Guidance": "Panduan Lapangan/pitch",
+ "Extract Features": "Ekstrak Fitur",
+
+ "Start Training": "mulai Training",
+ "Generate Index": "Menghasilkan Index",
+
+ "Voice Model": "Model Suara",
+ "Index File": "Berkas Indeks",
+ "Refresh": "Muat ulang",
+ "Unload Voice": "Bongkar Suara",
+
+ "Single": "Sendiri",
+ "Upload Audio": "Ungah Audio",
+ "Select Audio": "Pilh Audio",
+ "Advanced Settings": "Pengaturan lanjutan",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Hapus Output (Menghapus semua audio di assets/audio)",
+ "Custom Output Path": "Kustom Jalur keluaran",
+ "Output Path": "Jalur keluaran",
+ "Split Audio": "pisah audio",
+ "Autotune": "Autotune",
+ "Pitch": "Nada",
+ "If >=3: apply median filtering to the harvested pitch results. The value represents the filter radius and can reduce breathiness": "If >=3: terapkan pemfilteran median pada hasil pitch yang dipanen. Nilai tersebut mewakili radius filter dan dapat mengurangi sesak napas",
+ "Search Feature Ratio": "Rasio Fitur Pencarian",
+ "Pitch extraction algorithm": "Algoritma ekstraksi nada",
+ "Convert": "Convert",
+ "Export Audio": "Export Audio",
+
+ "Batch": "Batch",
+ "Input Folder": "Input berkas",
+ "Enter input path": "Masukkan jalur masukan",
+ "Output Folder": "Folder Keluaran",
+ "Enter output path": "Masukkan jalur keluaran",
+
+ "Get information about the audio": "Dapatkan informasi tentang audio",
+ "Information about the audio file": "Informasi tentang file audio",
+ "Waiting for information...": "Menunggu informasi...",
+
+ "Model fusion": "Penggabungan model",
+ "Weight for Model A": "Weight untuk model A",
+ "Whether the model has pitch guidance": "Apakah model memiliki panduan nada",
+ "Model architecture version": "Versi arsitektur model",
+ "Path to Model A": "Jalan Menuju Model A",
+ "Path to Model B": "Jalan Menuju Model B",
+ "Path to model": "Jalan Menuju Model",
+ "Model information to be placed": "Informasi model yang akan ditempatkan",
+ "Fusion": "Fusi",
+
+ "Modify model information": "Ubah informasi model",
+ "Path to Model": "Jalan Menuju Model",
+ "Model information to be modified": "Informasi model yang akan dimodifikasi",
+ "Save file name": "Simpan nama file",
+ "Modify": "Ubah",
+
+ "View model information": "tampilkan informasi model",
+ "View": "tampilakan",
+ "Model extraction": "Ekstraksi model",
+ "Model conversion": "Konversi model",
+ "Pth file": "file Pth",
+ "Output of the pth file": "Keluaran dari file pth",
+
+ "# How to Report an Issue on GitHub": "# Cara Melaporkan Masalah di GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Klik tombol 'Rekam Layar' di bawah untuk mulai merekam masalah yang Anda alami.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. pergi ke [GitHub Issues](https://github.com/IAHispano/Applio/issues) dan klik tombol 'Masalah Baru'.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Lengkapi templat masalah yang disediakan, pastikan untuk menyertakan detail sesuai kebutuhan, dan manfaatkan bagian aset untuk mengunggah file rekaman dari langkah sebelumnya.",
+
+ "Record Screen": "Rekam layar",
+ "Record": "Rekam",
+ "Stop Recording": "Berhenti merekam",
+
+ "Introduce the model .pth path": "Perkenalkan jalur model .pth",
+ "See Model Information": "Lihat informasi model",
+
+ "## Download Model": "## Unduh Model",
+ "Model Link": "Tautan model",
+ "Introduce the model link": "Perkenalkan tautan model",
+ "Download Model": "Unduh Model",
+ "## Drop files": "## masukkan file ",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Seret file .pth dan file .index ke dalam ruang ini. Seret yang pertama, lalu yang lainnya.",
+
+ "TTS Voices": "suara TTS",
+ "Text to Synthesize": "Text ke disintesis",
+ "Or you can upload a .txt file": "Atau Anda dapat mengunggah file .txt",
+ "Enter text to synthesize": "Masukkan teks untuk disintesis",
+ "Output Path for TTS Audio": "Jalur Keluaran untuk Audio TTS",
+ "Output Path for RVC Audio": "Jalur Keluaran untuk Audio RVC",
+
+ "Enable Applio integration with Discord presence": "Aktifkan integrasi Applio dengan kehadiran Discord",
+ "Enable Applio integration with applio.org/models using flask": "Aktifkan integrasi Applio dengan applio.org/models menggunakan flask",
+ "Theme": "Tema (Diperlukan mulai ulang)",
+
+ "Plugin Installer": "Penginstal Plugin",
+ "Drag your plugin.zip to install it": "Seret plugin.zip Anda untuk menginstalnya",
+
+ "Version Checker": "Pemeriksa versi",
+ "Check for updates": "Periksa pembaruan"
+}
diff --git a/assets/i18n/languages/it_IT.json b/assets/i18n/languages/it_IT.json
new file mode 100644
index 0000000000000000000000000000000000000000..3bd46972af4bbe775f39633977dd84ab207be865
--- /dev/null
+++ b/assets/i18n/languages/it_IT.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Strumento di clonazione vocale definitivo, meticolosamente ottimizzato per una potenza, una modularità e un'esperienza user-friendly senza rivali.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Questa sezione contiene alcune utilità aggiuntive che spesso possono essere in fase sperimentale.",
+ "Output Information": "Informazioni sull'output",
+ "The output information will be displayed here.": "Qui verranno visualizzate le informazioni sull'output.",
+ "Inference": "Inferenza",
+ "Train": "Treno",
+ "Extra": "Supplementare",
+ "Merge Audios": "Unisci audio",
+ "Processing": "Elaborazione",
+ "Audio Analyzer": "Analizzatore audio",
+ "Model Information": "Informazioni sul modello",
+ "Plugins": "Plugin",
+ "Download": "Scaricare",
+ "Report a Bug": "Segnala un bug",
+ "Settings": "Impostazioni",
+ "Preprocess": "Pre-processo",
+ "Model Name": "Nome del modello",
+ "Name of the new model.": "Nome del nuovo modello.",
+ "Enter model name": "Immettere il nome del modello",
+ "Dataset Path": "Percorso del set di dati",
+ "Path to the dataset folder.": "Percorso della cartella del set di dati.",
+ "Refresh Datasets": "Aggiornare i set di dati",
+ "Dataset Creator": "Creatore di set di dati",
+ "Dataset Name": "Nome set di dati",
+ "Name of the new dataset.": "Nome del nuovo set di dati.",
+ "Enter dataset name": "Immettere il nome del set di dati",
+ "Upload Audio Dataset": "Carica set di dati audio",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Il file audio è stato aggiunto correttamente al set di dati. Fare clic sul pulsante di pre-elaborazione.",
+ "Enter dataset path": "Immettere il percorso del set di dati",
+ "Sampling Rate": "Frequenza di campionamento",
+ "The sampling rate of the audio files.": "La frequenza di campionamento dei file audio.",
+ "RVC Version": "Versione RVC",
+ "The RVC version of the model.": "La versione RVC del modello.",
+ "Preprocess Dataset": "Set di dati di pre-elaborazione",
+ "Extract": "Estrarre",
+ "Hop Length": "Lunghezza del luppolo",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Indica il tempo necessario affinché il sistema passi a un cambiamento di intonazione significativo. Le lunghezze di hop più piccole richiedono più tempo per l'inferenza, ma tendono a produrre una maggiore precisione dell'intonazione.",
+ "Batch Size": "Dimensione del lotto",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Si consiglia di allinearlo con la VRAM disponibile della GPU. Un'impostazione di 4 offre una maggiore precisione ma un'elaborazione più lenta, mentre 8 fornisce risultati più rapidi e standard.",
+ "Save Every Epoch": "Salva ogni epoca",
+ "Determine at how many epochs the model will saved at.": "Determinare il numero di epoche in cui verrà salvato il modello.",
+ "Total Epoch": "Epoca totale",
+ "Specifies the overall quantity of epochs for the model training process.": "Specifica la quantità complessiva di epoche per il processo di training del modello.",
+ "Pretrained": "Pre-addestrato",
+ "Save Only Latest": "Salva solo le ultime notizie",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Abilitando questa impostazione, i file G e D salveranno solo le versioni più recenti, risparmiando spazio di archiviazione.",
+ "Save Every Weights": "Risparmia ogni peso",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Questa impostazione consente di salvare i pesi del modello alla fine di ogni epoca.",
+ "Custom Pretrained": "Pre-addestrato personalizzato",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "L'utilizzo di modelli pre-addestrati personalizzati può portare a risultati superiori, poiché la selezione dei modelli pre-addestrati più adatti al caso d'uso specifico può migliorare significativamente le prestazioni.",
+ "Upload Pretrained Model": "Carica modello pre-addestrato",
+ "Refresh Custom Pretraineds": "Aggiornare i pre-addestrati personalizzati",
+ "Pretrained Custom Settings": "Impostazioni personalizzate pre-addestrate",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Il file eliminato non è un file pre-addestrato valido. Si prega di riprovare.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Fare clic sul pulsante Aggiorna per visualizzare il file pre-addestrato nel menu a discesa.",
+ "Pretrained G Path": "G preaddestrato personalizzato",
+ "Pretrained D Path": "D pre-addestrato personalizzato",
+ "GPU Settings": "Impostazioni GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Imposta le impostazioni avanzate della GPU, consigliate per gli utenti con una migliore architettura GPU.",
+ "GPU Custom Settings": "Impostazioni personalizzate GPU",
+ "GPU Number": "Numero GPU",
+ "0 to ∞ separated by -": "Da 0 a ∞ separati da -",
+ "GPU Information": "Informazioni sulla GPU",
+ "Pitch Guidance": "Guida al passo",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Utilizzando la guida dell'intonazione, diventa possibile rispecchiare l'intonazione della voce originale, compresa la sua altezza. Questa funzione è particolarmente utile per il canto e altri scenari in cui è essenziale preservare la melodia o il modello di intonazione originale.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Utilizza modelli pre-addestrati quando addestri i tuoi. Questo approccio riduce la durata della formazione e migliora la qualità complessiva.",
+ "Extract Features": "Estrai feature",
+ "Start Training": "Inizia la formazione",
+ "Generate Index": "Genera indice",
+ "Voice Model": "Modello vocale",
+ "Select the voice model to use for the conversion.": "Selezionare il modello vocale da utilizzare per la conversione.",
+ "Index File": "File di indice",
+ "Select the index file to use for the conversion.": "Selezionare il file di indice da utilizzare per la conversione.",
+ "Refresh": "Aggiornare",
+ "Unload Voice": "Scarica voce",
+ "Single": "Singolo",
+ "Upload Audio": "Carica audio",
+ "Select Audio": "Seleziona Audio",
+ "Select the audio to convert.": "Seleziona l'audio da convertire.",
+ "Advanced Settings": "Impostazioni avanzate",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Cancella output (elimina tutti gli audio nelle risorse/audio)",
+ "Custom Output Path": "Percorso di output personalizzato",
+ "Output Path": "Percorso di output",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Il percorso in cui verrà salvato l'audio in uscita, per impostazione predefinita in assets/audios/output.wav",
+ "Split Audio": "Dividere l'audio",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Dividi l'audio in blocchi per l'inferenza per ottenere risultati migliori in alcuni casi.",
+ "Autotune": "Sintonizzazione automatica",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Applica un soft autotune alle tue inferenze, consigliato per le conversioni di canto.",
+ "Clean Audio": "Audio pulito",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Pulisci l'uscita audio utilizzando algoritmi di rilevamento del rumore, consigliati per l'audio parlato.",
+ "Clean Strength": "Forza pulita",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Imposta il livello di pulizia sull'audio che desideri, più lo aumenti più si pulirà, ma è possibile che l'audio sia più compresso.",
+ "Pitch": "Pece",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Imposta il tono dell'audio, più alto è il valore, più alto è il tono.",
+ "Filter Radius": "Raggio del filtro",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Se il numero è maggiore o uguale a tre, l'impiego del filtraggio mediano sui risultati del tono raccolto ha il potenziale per ridurre la respirazione.",
+ "Search Feature Ratio": "Rapporto feature di ricerca",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Influenza esercitata dal file indice; Un valore più alto corrisponde a una maggiore influenza. Tuttavia, optare per valori più bassi può aiutare a mitigare gli artefatti presenti nell'audio.",
+ "Volume Envelope": "Inviluppo del volume",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Sostituite o miscelate con l'inviluppo del volume dell'output. Più il rapporto è vicino a 1, più viene impiegato l'inviluppo di uscita.",
+ "Protect Voiceless Consonants": "Proteggi le consonanti sorde",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Salvaguarda le consonanti distinte e i suoni respiratori per prevenire lacerazioni elettroacustiche e altri artefatti. Portando il parametro al valore massimo di 0,5 si ottiene una protezione completa. Tuttavia, la riduzione di questo valore potrebbe ridurre l'estensione della protezione, riducendo al contempo l'effetto di indicizzazione.",
+ "Pitch extraction algorithm": "Algoritmo di estrazione del passo",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Algoritmo di estrazione dell'intonazione da utilizzare per la conversione audio. L'algoritmo predefinito è rmvpe, consigliato per la maggior parte dei casi.",
+ "Convert": "Convertire",
+ "Export Audio": "Esporta audio",
+ "Batch": "Lotto",
+ "Input Folder": "Cartella di input",
+ "Select the folder containing the audios to convert.": "Seleziona la cartella contenente gli audio da convertire.",
+ "Enter input path": "Immettere il percorso di input",
+ "Output Folder": "Cartella di output",
+ "Select the folder where the output audios will be saved.": "Seleziona la cartella in cui verranno salvati gli audio in uscita.",
+ "Enter output path": "Immettere il percorso di output",
+ "Get information about the audio": "Ottenere informazioni sull'audio",
+ "Information about the audio file": "Informazioni sul file audio",
+ "Waiting for information...": "In attesa di informazioni...",
+ "## Voice Blender": "## Frullatore vocale",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Seleziona due modelli vocali, imposta la percentuale di fusione desiderata e uniscili in una voce completamente nuova.",
+ "Voice Blender": "Frullatore vocale",
+ "Drag and drop your model here": "Trascina e rilascia qui il tuo modello",
+ "You can also use a custom path.": "È anche possibile utilizzare un percorso personalizzato.",
+ "Blend Ratio": "Rapporto di fusione",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Regolando la posizione più verso un lato o l'altro si renderà il modello più simile al primo o al secondo.",
+ "Fusion": "Fusione",
+ "Path to Model": "Percorso verso il modello",
+ "Enter path to model": "Immettere il percorso per il modello",
+ "Model information to be placed": "Informazioni sul modello da posizionare",
+ "Inroduce the model information": "Introdurre le informazioni sul modello",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Le informazioni da inserire nel modello (puoi lasciarle vuote o inserire qualsiasi cosa).",
+ "View model information": "Visualizzare le informazioni sul modello",
+ "Introduce the model pth path": "Introdurre il percorso pth del modello",
+ "View": "Vista",
+ "Model extraction": "Estrazione del modello",
+ "Model conversion": "Conversione del modello",
+ "Pth file": "File Pth",
+ "Output of the pth file": "Output del file pth",
+ "# How to Report an Issue on GitHub": "# Come segnalare un problema su GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Fare clic sul pulsante \"Registra schermo\" in basso per avviare la registrazione del problema riscontrato.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Una volta terminata la registrazione del problema, fai clic sul pulsante \"Interrompi registrazione\" (lo stesso pulsante, ma l'etichetta cambia a seconda che tu stia registrando attivamente o meno).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Vai su [GitHub Issues](https://github.com/IAHispano/Applio/issues) e fai clic sul pulsante \"Nuovo problema\".",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Completa il modello di problema fornito, assicurandoti di includere i dettagli necessari e utilizza la sezione delle risorse per caricare il file registrato dal passaggio precedente.",
+ "Record Screen": "Registra schermo",
+ "Record": "Registrazione",
+ "Stop Recording": "Interrompi registrazione",
+ "Introduce the model .pth path": "Introdurre il percorso .pth del modello",
+ "See Model Information": "Vedere le informazioni sul modello",
+ "## Download Model": "## Scarica il modello",
+ "Model Link": "Collegamento al modello",
+ "Introduce the model link": "Introdurre il collegamento al modello",
+ "Download Model": "Scarica il modello",
+ "## Drop files": "## Trascina i file",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Trascina il file .pth e il file .index in questo spazio. Trascina uno e poi l'altro.",
+ "TTS Voices": "Voci TTS",
+ "Select the TTS voice to use for the conversion.": "Selezionare la voce TTS da utilizzare per la conversione.",
+ "Text to Synthesize": "Testo da sintetizzare",
+ "Enter the text to synthesize.": "Immettere il testo da sintetizzare.",
+ "Or you can upload a .txt file": "In alternativa, è possibile caricare un file .txt",
+ "Enter text to synthesize": "Immettere il testo da sintetizzare",
+ "Output Path for TTS Audio": "Percorso di uscita per l'audio TTS",
+ "Output Path for RVC Audio": "Percorso di uscita per l'audio RVC",
+ "Enable Applio integration with Discord presence": "Abilita l'integrazione di Applio con la presenza di Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Attiverà la possibilità di visualizzare l'attuale attività di Applio in Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Abilita l'integrazione di Applio con applio.org/models utilizzando il pallone",
+ "It will activate the possibility of downloading models with a click from the website.": "Attiverà la possibilità di scaricare i modelli con un click dal sito web.",
+ "Theme": "Tema",
+ "Select the theme you want to use. (Requires restarting Applio)": "Selezionare il tema che si desidera utilizzare. (Richiede il riavvio di Applio)",
+ "Language": "Lingua",
+ "Select the language you want to use. (Requires restarting Applio)": "Selezionare la lingua che si desidera utilizzare. (Richiede il riavvio di Applio)",
+ "Plugin Installer": "Programma di installazione del plug-in",
+ "Drag your plugin.zip to install it": "Trascina il plugin.zip per installarlo",
+ "Version Checker": "Controllo della versione",
+ "Check which version of Applio is the latest to see if you need to update.": "Controlla quale versione di Applio è l'ultima per vedere se è necessario eseguire l'aggiornamento.",
+ "Check for updates": "Controlla gli aggiornamenti"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/ja_JA.json b/assets/i18n/languages/ja_JA.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b03e82cd1af637296161dae7ad83d6acfac4bb1
--- /dev/null
+++ b/assets/i18n/languages/ja_JA.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "他の追随を許さないパワー、モジュール性、そしてユーザーフレンドリーな操作性を実現するために綿密に最適化された究極のボイスクローニングツール。",
+ "This section contains some extra utilities that often may be in experimental phases.": "このセクションには、多くの場合、実験段階にある可能性のあるいくつかの追加のユーティリティが含まれています。",
+ "Output Information": "出力情報",
+ "The output information will be displayed here.": "出力情報がここに表示されます。",
+ "Inference": "推論",
+ "Train": "学習",
+ "Extra": "おまけ",
+ "Merge Audios": "オーディオをマージ (Merge Audios)",
+ "Processing": "加工",
+ "Audio Analyzer": "オーディオアナライザ",
+ "Model Information": "モデル情報",
+ "Plugins": "プラグイン",
+ "Download": "ダウンロード",
+ "Report a Bug": "バグを報告する",
+ "Settings": "設定",
+ "Preprocess": "前処理",
+ "Model Name": "モデル名",
+ "Name of the new model.": "新しいモデルの名前。",
+ "Enter model name": "モデル名を入力",
+ "Dataset Path": "データセット パス",
+ "Path to the dataset folder.": "データセット フォルダーへのパス。",
+ "Refresh Datasets": "データセットの更新",
+ "Dataset Creator": "データセットクリエーター",
+ "Dataset Name": "データセット名",
+ "Name of the new dataset.": "新しいデータセットの名前。",
+ "Enter dataset name": "データセット名を入力",
+ "Upload Audio Dataset": "オーディオデータセットのアップロード",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "オーディオ ファイルがデータセットに正常に追加されました。前処理ボタンをクリックしてください。",
+ "Enter dataset path": "データセットのパスを入力",
+ "Sampling Rate": "サンプリングレート",
+ "The sampling rate of the audio files.": "オーディオファイルのサンプリングレートです。",
+ "RVC Version": "RVC バージョン",
+ "The RVC version of the model.": "モデルの RVC バージョン。",
+ "Preprocess Dataset": "データセットの前処理",
+ "Extract": "抽出",
+ "Hop Length": "ホップ長(Hop Length)",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "システムがピッチを大きく変化させるまでの時間を示します。ホップ長が短いほど、推論に時間がかかりますが、ピッチの精度は高くなる傾向があります。",
+ "Batch Size": "バッチサイズ",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "GPUの使用可能なVRAMに合わせることをお勧めします。4 に設定すると精度は向上しますが処理は遅くなり、8 に設定すると、より高速で標準的な結果が得られます。",
+ "Save Every Epoch": "すべてのエポックを保存",
+ "Determine at how many epochs the model will saved at.": "モデルが何エポックごとに保存されるかを指定します。",
+ "Total Epoch": "総エポック",
+ "Specifies the overall quantity of epochs for the model training process.": "総エポック数を指定します。",
+ "Pretrained": "事前学習済みモデル",
+ "Save Only Latest": "最新のみ保存",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "この設定を有効にすると、GとDファイルの最新バージョンのみを保存し、ストレージ容量を効果的に節約します。",
+ "Save Every Weights": "すべてのウェイトを保存(Save Every Weights)",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "この設定により、各エポックの終了時にモデルの重みを保存できます。",
+ "Custom Pretrained": "カスタム事前学習済みモデル",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "カスタム事前学習済みモデルを利用することで、優れた結果を得ることができます。特定のユースケースに合わせて最も適した事前学習済みモデルを選択することで、パフォーマンスが大幅に向上する可能性があります。",
+ "Upload Pretrained Model": "事前学習済みモデルのアップロード",
+ "Refresh Custom Pretraineds": "カスタム事前学習済みモデルの更新",
+ "Pretrained Custom Settings": "事前学習済みモデルのカスタム設定",
+ "The file you dropped is not a valid pretrained file. Please try again.": "ドロップしたファイルは有効な事前学習済みモデルファイルではありません。もう一度やり直してください。",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "更新ボタンをクリックすると、ドロップダウンメニューに事前学習済みモデルが表示されます。",
+ "Pretrained G Path": "カスタム事前学習済みモデル G",
+ "Pretrained D Path": "カスタム事前学習済みモデル D",
+ "GPU Settings": "GPU 設定",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "GPUアーキテクチャが優れているユーザーに推奨される、高度なGPU設定を設定します。",
+ "GPU Custom Settings": "GPU カスタム設定",
+ "GPU Number": "GPU番号",
+ "0 to ∞ separated by -": "範囲は0 から ∞ で、-で区切られます。",
+ "GPU Information": "GPU 情報",
+ "Pitch Guidance": "ピッチガイダンス",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "ピッチガイダンスを採用することで、ピッチを含め、元の声のイントネーションを反映させることが可能になります。この機能は、歌唱など、オリジナルのメロディーやピッチパターンを保持することが不可欠な場面で特に有用です。",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "独自のトレーニングを行う場合は、事前学習済みモデルを利用します。このアプローチにより、学習時間が短縮され、全体的な品質が向上します。",
+ "Extract Features": "特徴量の抽出",
+ "Start Training": "トレーニングを開始",
+ "Generate Index": "インデックスの生成",
+ "Voice Model": "音声モデル",
+ "Select the voice model to use for the conversion.": "変換に使用する音声モデルを選択します。",
+ "Index File": "インデックス ファイル",
+ "Select the index file to use for the conversion.": "変換に使用するインデックスファイルを選択します。",
+ "Refresh": "リフレッシュ",
+ "Unload Voice": "モデルのアンロード",
+ "Single": "シングル",
+ "Upload Audio": "オーディオのアップロード",
+ "Select Audio": "オーディオを選択",
+ "Select the audio to convert.": "変換するオーディオを選択します。",
+ "Advanced Settings": "詳細設定",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Clear Outputs (アセット/オーディオ内のすべてのオーディオを削除します)",
+ "Custom Output Path": "カスタム出力パス",
+ "Output Path": "出力パス",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "出力オーディオが保存されるパス (デフォルトでは assets/audios/output.wav",
+ "Split Audio": "オーディオの分割",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "音声をチャンクに分割して推論すると、より良い結果が得られる場合があります。",
+ "Autotune": "オートチューン",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "推論にソフトオートチューンを適用します。歌の変換に推奨されます。",
+ "Clean Audio": "クリーンオーディオ",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "音声を話すのに推奨されるノイズ検出アルゴリズムを使用して、音声出力をクリーンアップします。",
+ "Clean Strength": "クリーンな強さ",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "クリーンアップレベルを必要なオーディオに設定すると、上げれば上げるほどクリーンアップされますが、オーディオがより圧縮される可能性があります。",
+ "Pitch": "ピッチ",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "オーディオのピッチを設定し、値が大きいほどピッチが高くなります。",
+ "Filter Radius": "フィルタ半径(Filter Radius)",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "数値が 3 以上の場合、収集されたトーン結果に中央値フィルタリングを使用すると、呼吸が減少する可能性があります。",
+ "Search Feature Ratio": "特徴量検索比率",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "インデックスファイルによって及ぼされる影響。値が大きいほど、インデックスの影響が大きくなります。ただし、低い値を選択すると、オーディオに存在するアーティファクトを軽減できます。",
+ "Volume Envelope": "ボリュームエンベロープ",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "出力のボリュームエンベロープで代用またはブレンドします。比率が 1 に近づくほど、出力エンベロープが採用されます。",
+ "Protect Voiceless Consonants": "無声子音の保護",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "明瞭な子音と呼吸音を保護し、電気音響の引き裂きやその他のアーチファクトを防ぎます。パラメータを最大値の 0.5 までプルすると、包括的な保護が提供されます。ただし、この値を小さくすると、インデックス作成の影響が軽減される可能性がある一方で、保護の範囲が狭くなる可能性があります。",
+ "Pitch extraction algorithm": "ピッチ抽出アルゴリズム",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "オーディオ変換に使用するピッチ抽出アルゴリズム。デフォルトのアルゴリズムは rmvpe で、ほとんどの場合に推奨されます。",
+ "Convert": "コンバート",
+ "Export Audio": "オーディオのエクスポート",
+ "Batch": "バッチ",
+ "Input Folder": "入力フォルダ",
+ "Select the folder containing the audios to convert.": "変換するオーディオを含むフォルダを選択します。",
+ "Enter input path": "入力パスを入力",
+ "Output Folder": "出力フォルダ",
+ "Select the folder where the output audios will be saved.": "出力オーディオを保存するフォルダを選択します。",
+ "Enter output path": "出力パスを入力",
+ "Get information about the audio": "オーディオに関する情報を取得する",
+ "Information about the audio file": "オーディオファイルに関する情報",
+ "Waiting for information...": "情報を待っています...",
+ "## Voice Blender": "## モデルマージ",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "2つの音声モデルを選択し、希望のブレンド率を設定して、まったく新しい音声モデルにマージします。",
+ "Voice Blender": "モデルマージ",
+ "Drag and drop your model here": "ここにモデルをドラッグ&ドロップします",
+ "You can also use a custom path.": "カスタムパスを使用することもできます。",
+ "Blend Ratio": "ブレンド比(Blend Ratio)",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "位置をどちらか一方に調整すると、モデルが最初または2番目に近づきます。",
+ "Fusion": "マージ",
+ "Path to Model": "モデルへのパス",
+ "Enter path to model": "モデルへのパスを入力",
+ "Model information to be placed": "配置するモデル情報",
+ "Inroduce the model information": "モデル情報の取得",
+ "The information to be placed in the model (You can leave it blank or put anything).": "モデルに配置する情報(空白のままでも、何でも入力できます)。",
+ "View model information": "モデル情報の表示",
+ "Introduce the model pth path": ".pthのパスを指定してください。",
+ "View": "表示",
+ "Model extraction": "モデルの抽出",
+ "Model conversion": "モデル変換",
+ "Pth file": "Pth ファイル",
+ "Output of the pth file": "p番目のファイルの出力",
+ "# How to Report an Issue on GitHub": "# GitHub で問題を報告する方法",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1.下の[画面の記録]ボタンをクリックして、発生している問題の記録を開始します。",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2.問題の記録が終了したら、[記録の停止]ボタンをクリックします(同じボタンですが、アクティブに記録しているかどうかによってラベルが変わります)。",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub Issues](https://github.com/IAHispano/Applio/issues)に移動し、[New Issue]ボタンをクリックします。",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. 提供された課題テンプレートに記入し、必要に応じて詳細を含め、アセット セクションを使用して前の手順で記録したファイルをアップロードします。",
+ "Record Screen": "録画画面",
+ "Record": "記録",
+ "Stop Recording": "記録の停止",
+ "Introduce the model .pth path": "モデルの .pth パスを導入する",
+ "See Model Information": "「モデル情報」を参照",
+ "## Download Model": "## モデルのダウンロード",
+ "Model Link": "モデルリンク",
+ "Introduce the model link": "モデルリンクの紹介",
+ "Download Model": "モデルのダウンロード",
+ "## Drop files": "## ファイルのドロップ",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": ".pth ファイルと .index ファイルをこのスペースにドラッグします。一方をドラッグしてから、もう一方をドラッグします。",
+ "TTS Voices": "TTS ボイス",
+ "Select the TTS voice to use for the conversion.": "変換に使用する TTS 音声を選択します。",
+ "Text to Synthesize": "合成するテキスト",
+ "Enter the text to synthesize.": "合成するテキストを入力します。",
+ "Or you can upload a .txt file": "または、.txtファイルをアップロードすることもできます",
+ "Enter text to synthesize": "合成するテキストを入力する",
+ "Output Path for TTS Audio": "TTSオーディオの出力パス",
+ "Output Path for RVC Audio": "RVCオーディオの出力パス",
+ "Enable Applio integration with Discord presence": "Applio と Discord のプレゼンスの統合を有効にする",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "これにより、Discordで現在のApplioアクティビティを表示する可能性がアクティブになります。",
+ "Enable Applio integration with applio.org/models using flask": "フラスコを使用して Applio と applio.org/models の統合を有効にする",
+ "It will activate the possibility of downloading models with a click from the website.": "ウェブサイトからクリックするだけでモデルをダウンロードする可能性がアクティブになります。",
+ "Theme": "テーマ",
+ "Select the theme you want to use. (Requires restarting Applio)": "使用するテーマを選択します。(Applioの再起動が必要)",
+ "Language": "言語",
+ "Select the language you want to use. (Requires restarting Applio)": "使用する言語を選択します。(Applioの再起動が必要)",
+ "Plugin Installer": "プラグインインストーラ",
+ "Drag your plugin.zip to install it": "plugin.zipをドラッグしてインストールします",
+ "Version Checker": "バージョンチェッカー",
+ "Check which version of Applio is the latest to see if you need to update.": "更新が必要かどうか、Applioのどのバージョンが最新かを確認してください。",
+ "Check for updates": "アップデートを確認する"
+}
diff --git a/assets/i18n/languages/jv_JV.json b/assets/i18n/languages/jv_JV.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e26dfeeb6e641a33dae4961196235bdb965b21b
--- /dev/null
+++ b/assets/i18n/languages/jv_JV.json
@@ -0,0 +1 @@
+{}
\ No newline at end of file
diff --git a/assets/i18n/languages/ko_KO.json b/assets/i18n/languages/ko_KO.json
new file mode 100644
index 0000000000000000000000000000000000000000..a2b25326b6e469414fd16d807b5beb87b5f1c95a
--- /dev/null
+++ b/assets/i18n/languages/ko_KO.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "타의 추종을 불허하는 성능, 모듈성 및 사용자 친화적인 경험을 위해 세심하게 최적화된 궁극의 음성 복제 도구입니다.",
+ "This section contains some extra utilities that often may be in experimental phases.": "이 섹션에는 아직 실험 단계에 있는 추가 유틸리티가 포함되어 있습니다.",
+ "Output Information": "출력 정보",
+ "The output information will be displayed here.": "출력 정보가 여기에 표시됩니다.",
+ "Inference": "추론",
+ "Train": "모델 학습",
+ "Extra": "기타 도구",
+ "Merge Audios": "오디오 병합",
+ "Processing": "처리",
+ "Audio Analyzer": "오디오 분석기",
+ "Model Information": "모델 정보",
+ "Plugins": "플러그인",
+ "Download": "다운로드",
+ "Report a Bug": "버그 신고",
+ "Settings": "설정",
+ "Preprocess": "사전 처리",
+ "Model Name": "모델명",
+ "Name of the new model.": "새 모델의 이름입니다.",
+ "Enter model name": "모델명 입력",
+ "Dataset Path": "데이터 세트 경로",
+ "Path to the dataset folder.": "데이터 세트 폴더의 경로입니다.",
+ "Refresh Datasets": "데이터 세트 새로 고침",
+ "Dataset Creator": "데이터셋 생성기",
+ "Dataset Name": "데이터 세트 이름",
+ "Name of the new dataset.": "새 데이터 세트의 이름입니다.",
+ "Enter dataset name": "데이터 세트 이름 입력",
+ "Upload Audio Dataset": "오디오 데이터 세트 업로드",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "오디오 파일이 데이터 세트에 성공적으로 추가되었습니다. 전처리 버튼을 클릭해주세요.",
+ "Enter dataset path": "데이터 세트 경로 입력",
+ "Sampling Rate": "샘플링 레이트",
+ "The sampling rate of the audio files.": "오디오 파일의 샘플링 레이트입니다.",
+ "RVC Version": "RVC 버전",
+ "The RVC version of the model.": "모델의 RVC 버전입니다.",
+ "Preprocess Dataset": "데이터 세트 사전 처리",
+ "Extract": "추출물",
+ "Hop Length": "홉 길이",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "시스템이 중요한 피치 변화로 전환되는 데 걸리는 시간을 나타냅니다. 홉 길이가 작을수록 추론에 더 많은 시간이 필요하지만 피치 정확도가 높아지는 경향이 있습니다.",
+ "Batch Size": "배치 크기(Batch Size)",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "GPU의 사용 가능한 VRAM에 맞추는 것이 좋습니다. 4로 설정하면 정확도가 향상되지만 처리 속도가 느려지고, 8로 설정하면 더 빠르고 표준적인 결과를 얻을 수 있습니다.",
+ "Save Every Epoch": "모든 Epoch를 저장합니다",
+ "Determine at how many epochs the model will saved at.": "모델이 몇 epoch에 저장될지 결정합니다.",
+ "Total Epoch": "Total Epoch(총 에포크)",
+ "Specifies the overall quantity of epochs for the model training process.": "모델 학습 프로세스에 대한 전체 epoch 수를 지정합니다.",
+ "Pretrained": "사전학습 모델 사용",
+ "Save Only Latest": "마지막 파일만 저장(Save Only Latest)",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "이 설정을 활성화하면 마지막으로 저장된 G 및 D 파일만 남게 되므로 저장 공간을 효과적으로 절약할 수 있습니다.",
+ "Save Every Weights": "모든 가중치 저장",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "이 설정을 사용하면 각 Epoch가 끝날 때 모델의 가중치를 저장할 수 있습니다.",
+ "Custom Pretrained": "커스텀 사전학습 모델 사용",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "특정 사용 사례에 맞는 가장 적합한 커스텀 사전학습 모델을 선택하면 성능이 크게 향상될 수 있고 우수한 결과를 얻을 수 있습니다.",
+ "Upload Pretrained Model": "사전학습 된 모델 업로드",
+ "Refresh Custom Pretraineds": "커스텀 사전학습 모델 새로 고침",
+ "Pretrained Custom Settings": "커스텀 사전학습 모델 설정",
+ "The file you dropped is not a valid pretrained file. Please try again.": "드롭한 파일이 유효한 사전학습 모델 파일이 아닙니다. 다시 시도하십시오.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "새로 고침 버튼을 클릭하면 드롭다운 메뉴에서 사전 학습 된 파일을 볼 수 있습니다.",
+ "Pretrained G Path": "커스텀 사전학습 모델 G",
+ "Pretrained D Path": "커스텀 사전학습 모델 D",
+ "GPU Settings": "GPU 설정",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "고급 GPU 아키텍처를 사용하는 사용자에게 권장되는 GPU 설정.",
+ "GPU Custom Settings": "GPU 커스텀 설정",
+ "GPU Number": "GPU 번호",
+ "0 to ∞ separated by -": "0 - ∞ - 로 구분",
+ "GPU Information": "GPU 정보",
+ "Pitch Guidance": "음정(Pitch) 안내",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "음정(Pitch) 가이던스를 사용하면 음정을 포함하여 원래 목소리의 억양을 미러링할 수 있습니다. 이 기능은 노래 및 원래 멜로디 또는 음정 패턴의 보존이 필수적인 것에 특히 유용합니다.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "자체 모델을 학습할 때 사전 학습된 모델을 활용합니다. 이 접근 방식은 학습 시간을 줄이고 전반적인 품질을 향상시킵니다.",
+ "Extract Features": "특징 추출",
+ "Start Training": "트레이닝 시작하기",
+ "Generate Index": "인덱스 생성",
+ "Voice Model": "음성 모델",
+ "Select the voice model to use for the conversion.": "변환에 사용할 음성 모델을 선택합니다.",
+ "Index File": "색인 파일",
+ "Select the index file to use for the conversion.": "변환에 사용할 인덱스 파일을 선택합니다.",
+ "Refresh": "새로고침",
+ "Unload Voice": "음성 언로드",
+ "Single": "싱글",
+ "Upload Audio": "오디오 업로드",
+ "Select Audio": "오디오 선택",
+ "Select the audio to convert.": "변환할 오디오를 선택합니다.",
+ "Advanced Settings": "고급 설정",
+ "Clear Outputs (Deletes all audios in assets/audios)": "출력 지우기(에셋/오디오의 모든 오디오 삭제)",
+ "Custom Output Path": "사용자 지정 출력 경로",
+ "Output Path": "출력 경로",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "출력 오디오가 저장될 경로 입니다, 기본값 assets/audios/output.wav",
+ "Split Audio": "오디오 분할",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "경우에 따라 더 나은 결과를 얻기 위해 오디오를 청크로 분할합니다.",
+ "Autotune": "오토튠",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "추론 시 소프트 오토튠을 적용합니다. 노래에 권장합니다",
+ "Clean Audio": "디-노이즈 오디오",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "노이즈 감지 알고리즘을 사용하여 출력되는 오디오의 불필요한 정보를 정리합니다. 연설과 같은 스피치 오디오에 적합합니다.",
+ "Clean Strength": "감지 강도",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "감지 레벨을 높이면 높일수록 정리가 더 많이 되지만 오디오가 더 압축되어 음질 저하가 생길 수 있습니다.",
+ "Pitch": "음정",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "오디오의 음정(Pitch)을 설정합니다. 값이 높을수록 음정이 높아집니다.",
+ "Filter Radius": "필터 반경",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "숫자가 3보다 크거나 같을 때 수집된 톤 결과에 중앙값 필터링을 사용하여 호흡등의 소리를 감소 시킬 수 있습니다",
+ "Search Feature Ratio": "특성 검색 비율(Search Feature Ratio)",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "특성 인덱스 파일이 미치는 영향; 값이 높을수록 모델의 특성을 잘 나타내지만 값이 낮을수록 인공적인 느낌을 줄어들 수도 있습니다.",
+ "Volume Envelope": "볼륨 엔벨로프(Volume Envelope)",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "출력의 볼륨 포락선(Envelope)로 대체하거나 혼합합니다. 비율이 1에 가까울수록 출력 엔벨로프가 더 많이 사용됩니다.",
+ "Protect Voiceless Consonants": "무성 자음 보호",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "뚜렷한 자음과 숨소리를 보호하여 전자 음향 찢어짐 및 기타 아티팩트를 방지합니다. 설정값을 최대치인 0.5로 끌어오면 포괄적인 보호가 제공됩니다. 반대로 보호 범위를 낮출 경우 인덱싱 효과를 완화할 수 있습니다",
+ "Pitch extraction algorithm": "음정(Pitch) 추출 알고리즘",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "오디오 변환에 사용할 음정 추출 알고리즘입니다. 디폴트 알고리즘은 rmvpe이며, 대부분의 경우에 권장됩니다.",
+ "Convert": "변환",
+ "Export Audio": "오디오 내보내기",
+ "Batch": "일괄",
+ "Input Folder": "입력 폴더",
+ "Select the folder containing the audios to convert.": "변환할 오디오가 포함된 폴더를 선택합니다.",
+ "Enter input path": "입력 경로 입력",
+ "Output Folder": "출력 폴더",
+ "Select the folder where the output audios will be saved.": "출력 오디오를 저장할 폴더를 선택합니다.",
+ "Enter output path": "출력 경로 입력",
+ "Get information about the audio": "오디오에 대한 정보 가져오기",
+ "Information about the audio file": "오디오 파일에 대한 정보",
+ "Waiting for information...": "정보를 기다리는 중...",
+ "## Voice Blender": "## 목소리 혼합",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "두 개의 음성 모델을 선택하고 원하는 혼합 비율을 설정하면 완전히 새로운 음성으로 혼합됩니다.",
+ "Voice Blender": "목소리 혼합",
+ "Drag and drop your model here": "여기에 모델을 끌어다 놓습니다.",
+ "You can also use a custom path.": "사용자 지정 경로를 사용할 수도 있습니다.",
+ "Blend Ratio": "합성 비율(Blend Ratio)",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "위치를 양쪽 방향으로 조정하며 두 모델의 혼합 비중을 선택할 수 있습니다.",
+ "Fusion": "목소리 혼합",
+ "Path to Model": "모델 경로",
+ "Enter path to model": "모델 경로 입력",
+ "Model information to be placed": "배치할 모델 정보",
+ "Inroduce the model information": "모델 정보 도입",
+ "The information to be placed in the model (You can leave it blank or put anything).": "모델의 정보(비워 두거나 아무 내용을 써도 됩니다).",
+ "View model information": "모델 정보 보기",
+ "Introduce the model pth path": "모델 pth 경로 소개",
+ "View": "보기",
+ "Model extraction": "모델 추출",
+ "Model conversion": "모델 변환",
+ "Pth file": "Pth 파일",
+ "Output of the pth file": "pth 파일의 출력",
+ "# How to Report an Issue on GitHub": "# GitHub에서 문제를 보고하는 방법",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. 아래의 '녹화 화면' 버튼을 클릭하여 발생한 문제의 녹화를 시작합니다.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. 문제 기록이 끝나면 '기록 중지' 버튼(동일한 버튼이지만 현재 기록 중인지 여부에 따라 레이블이 변경됨)을 클릭합니다.",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub 문제](https://github.com/IAHispano/Applio/issues)로 이동하여 '새 문제' 버튼을 클릭합니다.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. 제공된 문제 템플릿을 완료하고 필요에 따라 세부 정보를 포함하고 자산 섹션을 활용하여 이전 단계에서 기록된 파일을 업로드합니다.",
+ "Record Screen": "녹화 화면",
+ "Record": "기록",
+ "Stop Recording": "기록 중지",
+ "Introduce the model .pth path": "모델 .pth 경로 소개",
+ "See Model Information": "모델 정보 보기",
+ "## Download Model": "## 모델 다운로드",
+ "Model Link": "모델 링크",
+ "Introduce the model link": "모델 소개 링크",
+ "Download Model": "모델 다운로드",
+ "## Drop files": "## 파일 드롭",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": ".pth 파일과 .index 파일을 이 공간으로 드래그합니다. 하나를 드래그한 다음 다른 하나를 드래그합니다.",
+ "TTS Voices": "TTS 음성",
+ "Select the TTS voice to use for the conversion.": "변환에 사용할 TTS 음성을 선택합니다.",
+ "Text to Synthesize": "합성할 텍스트(Text to Synthesize)",
+ "Enter the text to synthesize.": "합성할 텍스트를 입력합니다.",
+ "Or you can upload a .txt file": "또는 .txt 파일을 업로드할 수 있습니다.",
+ "Enter text to synthesize": "합성할 텍스트 입력",
+ "Output Path for TTS Audio": "TTS 오디오의 출력 경로",
+ "Output Path for RVC Audio": "RVC 오디오의 출력 경로",
+ "Enable Applio integration with Discord presence": "Discord와 Applio 통합 활성화",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Discord에서 Applio의 활동이 표시되는 기능을 활성화합니다.",
+ "Enable Applio integration with applio.org/models using flask": "플라스크를 사용하여 applio.org/models 와 Applio 통합 활성화",
+ "It will activate the possibility of downloading models with a click from the website.": "웹 사이트에서 클릭 한 번으로 모델을 다운로드 할 수 있습니다.",
+ "Theme": "테마",
+ "Select the theme you want to use. (Requires restarting Applio)": "사용할 테마를 선택합니다. (Applio를 다시 시작해야 함)",
+ "Language": "언어",
+ "Select the language you want to use. (Requires restarting Applio)": "사용할 언어를 선택합니다. (Applio를 다시 시작해야 함)",
+ "Plugin Installer": "플러그인 인스톨러",
+ "Drag your plugin.zip to install it": "plugin.zip 드래그하여 설치하십시오.",
+ "Version Checker": "버전 검사기",
+ "Check which version of Applio is the latest to see if you need to update.": "Applio의 최신 버전을 확인하여 업데이트가 필요한지 확인하십시오.",
+ "Check for updates": "업데이트 확인"
+}
diff --git a/assets/i18n/languages/ml_IN.json b/assets/i18n/languages/ml_IN.json
new file mode 100644
index 0000000000000000000000000000000000000000..d20de22513a65792558614f31649db13f9f95620
--- /dev/null
+++ b/assets/i18n/languages/ml_IN.json
@@ -0,0 +1,204 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "അനന്തമായ ശക്തി, മോഡുലാരിറ്റി, ഉപയോക്തൃ-സൗഹൃദ അനുഭവത്തിനായി യാഥാർത്ഥ്യമാക്കിയ അത്യുന്നതമായ വോയ്സ് ക്ലോണിങ് ടൂൾ.\n[മലയാളത്തിലേക്ക് വായന: Enes](https://discord.com/users/1140031358006202468)",
+ "This section contains some extra utilities that often may be in experimental phases.": "ഈ പേരിലെ കൊണ്ടാടാൻ പ്രയോജനപ്രദമായ എന്നതിനാൽ കഴിഞ്ഞത് സാധാരണയായ പോസിക്കുകളിൽ അവസാനിക്കാത്ത ചില യന്ത്രങ്ങൾ ഉള്ളതാണ്.",
+ "Output Information": "പ്രണാമം വിവരം",
+ "The output information will be displayed here.": "ഇവിടെ പ്രണയ വിവരങ്ങൾ പ്രദശിപ്പിക്കപ്പെടും.",
+ "Inference": "സൂചന",
+ "Train": "പ്രശിക്ഷണം",
+ "Extra": "അധികം",
+ "Merge Audios": "ഓഡിയോ ഒടിക്കുക",
+ "Processing": "പ്രൊസസ്സിംഗ്",
+ "Audio Analyzer": "ഓഡിയോ വിശ്ലേഷണകൾ",
+ "Model Information": "മോഡൽ വിവരങ്ങൾ",
+ "Plugins": "പ്ലഗിൻസ്",
+ "Download": "ഡൗൺലോഡ്",
+ "Report a Bug": "പിശക് റിപ്പോർട്ട്",
+ "Settings": "സെറ്റിംഗുകൾ",
+ "Preprocess": "പ്രൊസസ്",
+ "Model Name": "മോഡൽ പേര്",
+ "Name of the new model.": "പുതിയ മോഡലിന്റെ പേര്.",
+ "Enter model name": "മോഡൽ പേര് നൽകുക",
+ "Dataset Path": "ഡാറ്റാസെറ്റ് പാത",
+ "Path to the dataset folder.": "ഡാറ്റാസെറ്റ് ഫോൾഡർക്കുള്ള പാത.",
+ "Refresh Datasets": "ഡാറ്റാസെറ്റുകൾ പുനഃസൃഷ്ടിക്കുക",
+ "Dataset Creator": "ഡാറ്റാസെറ്റ് സൃഷ്ടാവ്",
+ "Dataset Name": "ഡാറ്റാസെറ്റ് പേര്",
+ "Name of the new dataset.": "പുതിയ ഡാറ്റാസെറ്റിന്റെ പേര്.",
+ "Enter dataset name": "ഡാറ്റാസെറ്റ് പേര് നൽകുക",
+ "Upload Audio Dataset": "ഓഡിയോ ഡാറ്റാസെറ്റ് അപ്ലോഡ് ചെയ്യുക",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ഓഡിയോ ഫയൽ യഥാർത്ഥമായി ഡാറ്റാസെറ്റിലേക്ക് ചേർന്നു. ദയവായി പ്രൊസെസ് ബട്ടൺ അമർത്തുക.",
+ "Enter dataset path": "ഡാറ്റാസെറ്റ് പാത നൽകുക",
+ "Sampling Rate": "സാമ്പ്ലിംഗ് റേറ്റ്",
+ "The sampling rate of the audio files.": "ഓഡിയോ ഫയലുകളുടെ സാമ്പ്ലിംഗ് റേറ്റ്.",
+ "RVC Version": "RVC പതിപ്പ്",
+ "The RVC version of the model.": "മോഡലിന്റെ RVC പതിപ്പ്.",
+ "Preprocess Dataset": "ഡാറ്റാസെറ്റ് പ്രൊസെസ് ചെയ്യുക",
+
+ "Embedder Model": "Embedder Model",
+ "Model used for learning speaker embedding.": "സ്പീക്കർ എംബെഡ്ഡിംഗ് പഠിപ്പിക്കാൻ ഉപയോഗിക്കുന്ന മോഡൽ.",
+ "Extract": "എക്സ്ട്രാക്ട്",
+ "Hop Length": "ഹോപ്പ് ലെന്ത്",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "സിസ്റ്റത്തിൽ ഒരു വിശദ പിച്ച് മാറ്റത്തിന്റെ ദാരിദ്ര്യം സൂചിപ്പിക്കുന്നു. ചെറിയ ഹോപ്പ് ലെന്തുകള് ഇൻഫരൻസിനായി കൂടുതൽ സമയം ആവശ്യപ്പെടുന്നു, എന്നിരുന്നാലും ഉയരമായ പിച്ച് അക്ക്യൂറസി നൽകുന്നു.",
+ "Batch Size": "ബാച്ച് വലുപ്പം",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "നിനക്ക് ലഭ്യമായ GPU-യുടെ VRAM നുസരിച്ച് അലൈൻ ചെയ്യുന്നത് പ്രാധാന്യപ്പെടുന്നു. 4-ന്റെ സെറ്റിംഗ് മൊത്തം അക്ക്യൂറസി പരിഷ്കർത്തനവും എല്ലാത്തിനെയും അവലംബപ്പെടുന്നു, എന്നിരുന്നാലും 8 വലുപ്പം അനുഭവജീവനത്തിനായി ഉന്നയിക്കുന്നു.",
+ "Save Every Epoch": "എന്നാൽ എന്റെ എപ്പൊക്കാലിലെയും മോഡൽ സേവ് ചെയ്യുന്നതിനു വരെ",
+ "Determine at how many epochs the model will be saved at.": "എന്താണ് എപ്പൊക്കാലിൽ മോഡൽ സേവ് ചെയ്യപ്പെടുന്നതെന്ന് തിരഞ്ഞെടുക്കുക.",
+ "Total Epoch": "മൊത്തം എപ്പൊക്ക്",
+ "Specifies the overall quantity of epochs for the model training process.": "മോഡൽ പ്രശിക്ഷണ പ്രക്രിയയ്ക്കായി എപ്പൊക്ക് എത്രയാണ് എന്നത് നിര്ദിഷ്ടമാക്കുന്നു.",
+ "Pretrained": "പ്രീട്രെയിനെഡ്",
+ "Save Only Latest": "കേവലായി പുതിയത് മാത്രം സേവ് ചെയ്യുക",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "ഈ സെറ്റിംഗ് പ്രവർത്തനത്തിനായി ക്രമീകരിച്ചാൽ, G മറ്റും D ഫയലുകൾ അവരുടെ അവസാന പതിപ്പുകൾക്ക് മാത്രം സേവ് ചെയ്യും, പ്രഭാവകരമായി സ്റ്റോറേജ് സ്ഥലം സംരക്ഷിക്കുന്നതാണ്.",
+ "Save Every Weights": "എന്റെ എപ്പൊക്കാലിലും ഭാരം സേവ് ചെയ്യുക",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "ഈ സെറ്റിംഗ് നിര്ദ്ദേശപ്പെടുന്ന ഓരോ എപ്പൊക്കിലും മോഡലിന്റെ ഭാരങ്ങൾ സേവ് ചെയ്യാൻ കഴിയുന്നു.",
+ "Custom Pretrained": "കസ്റ്റം പ്രീട്രെയിനെഡ്",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "കസ്റ്റം പ്രീട്രെയിനെഡ് മോഡലുകൾ ഉപയോഗിക്കുന്നത് ഉന്നത ഫലങ്ങൾ നൽകിയാൽ, സ്പഷ്ടമായ ഉപയോഗ കേസിനായി കൈമേലെ പ്രീട്രെയിനെഡ് മോഡലുകൾ തിരഞ്ഞെടുക്കുന്നത് സാധാരണയായ പ്രദർശനത്തെ വളർത്തുന്നതിനും പ്രദർശനം വളർത്തുന്നതിനും കൂടുതൽ സഹായകമാകും.",
+ "Upload Pretrained Model": "പ്രീട്രെയിനെഡ് മോഡൽ അപ്ലോഡ് ചെയ്യുക",
+ "Refresh Custom Pretraineds": "കസ്റ്റം പ്രീട്രെയിനെഡുകൾ പുനഃസ്വന്തമാക്കുക",
+ "Pretrained Custom Settings": "പ്രീട്രെയിനെഡ് കസ്റ്റം സെറ്റിംഗുകൾ",
+ "The file you dropped is not a valid pretrained file. Please try again.": "നിനക്ക് ഡ്രോപ്പ് ചെയ്യിയ ഫയൽ ഒരു സാധാരണ പ്രീട്രെയിനെഡ് ഫയലല്ല. ദയവായി വീണ്ടും ശ്രയിക്കുക.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "പ്രീട്രെയിനെഡ് ഫയലെ ഡ്രോപ്പ്ഡൌൺ മെനുവിലെത്താൻ റെഫ്രഷ് ബട്ടൺ ക്ലിക്കുചെയ്യുക.",
+ "Pretrained G Path": "കസ്റ്റം പ്രീട്രെയിനെഡ് G പാത",
+ "Pretrained D Path": "കസ്റ്റം പ്രീട്രെയിനെഡ് D പാത",
+ "GPU Settings": "GPU സെറ്റിംഗുകൾ",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "അട്വാൻസ്ഡ് GPU സെറ്റിംഗുകൾ സജ്ജീവമാക്കുന്നു, പ്രശസ്ത GPU ആർക്കിടെയുള്ള ഉപയോക്താക്കളിനായി ശിഫാരസ് ചെയ്തത്.",
+ "GPU Custom Settings": "GPU കസ്റ്റം സെറ്റിംഗുകൾ",
+ "GPU Number": "GPU നമ്പർ",
+ "0 to ∞ separated by -": "0 മുതൽ ∞ വരെ - ഒടുക്കിയ",
+ "GPU Information": "GPU വിവരം",
+ "Pitch Guidance": "പിച്ച് ഗൈഡൻസ്",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "പിച്ച് ഗൈഡൻസ് ഉപയോഗിച്ച്, ഓരിജിനൽ വോയ്സിന്റെ ഇൻറോണേഷനെ, അതിന്റെ പിച്ചു സേവനേന്ന്, സോണ്റിംഗ് മുതലായ സന്നിധികളിൽ പ്രാഥമിക സ്വരം അല്ലെങ്കിൽ പിച്ച് നമ്പറെ പരിരക്ഷിക്കേണ്ടതായ സ്ഥിതികളിലേക്ക് മികച്ച ഫലങ്ങൾ പ്രാപ്തമാക്കാനാകുന്നു.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "നിനക്ക് സ്വന്തമായി പ്രശിക്ഷണം നടത്തുമ്പോൾ പ്രീട്രെയിനെഡ് മോഡലുകൾ ഉപയോഗിക്കുക. ഈ രീതി പ്രശിക്ഷണ സമയം കുറയ്ക്കുന്നുവെങ്കിൽ മൊത്തം ഗുണമേന്മ വരും.",
+ "Extract Features": "ഫീച്ചർ എക്സ്ട്രാക്ട്",
+ "Overtraining Detector": "ഓവർട്രെയിനിംഗ് ഡിറ്റക്റ്റർ",
+ "Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data.": "പ്രശിക്ഷണ ഡാറ്റയെ നിങ്ങളുടെ മോഡൽ അതിന്റെ തരംതിരിച്ചു പഠിക്കാൻ അനുവദിക്കുന്നത് നിലവിൽ നിന്ന് ഓവർട്രെയിനിംഗ് ശ്രമിക്കുക.",
+ "Overtraining Detector Settings": "ഓവർട്രെയിനിംഗ് ഡിറ്റക്റ്റർ സെറ്റിംഗുകൾ",
+ "Overtraining Threshold": "ഓവർട്രെയിനിംഗ് താഴ്ന്ന മിതം",
+ "Set the maximum number of epochs you want your model to stop training if no improvement is detected.": "യാത്രാവധി പരിശോധിച്ചിട്ടില്ലാത്ത അഭിവൃദ്ധി നിരയെടുക്കുകയാണെങ്കിൽ നിങ്ങളുടെ മോഡൽ പരിശോധനയെന്നത് പ്രശിക്ഷണം നിലനിൽക്കാനുള്ള ഏറ്റവും ഉന്നത എപ്പോക്കുകൾ സജ്ജമാക്കുക.",
+
+ "Start Training": "പ്രശിക്ഷണം ആരംഭിക്കുക",
+ "Stop Training & Restart Applio": "പ്രശിക്ഷണം നിർത്തുക & അപ്ലിയോ പുനഃപ്രാരംഭിക്കുക",
+ "Generate Index": "ഇൻഡെക്സ് സൃഷ്ടിക്കുക",
+
+ "Export Model": "എക്സ്പോർട്ട് മോഡൽ",
+ "The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive.": "'അപ്ലോഡ്' ബട്ടൺ കേവലം ഗൂഗിൾ കോളാബിന് മാത്രം: നിങ്ങളുടെ ഗൂഗിൾ ഡ്രൈവിലെ ApplioExported ഫോൾഡറിലേക്ക് എക്സ്പോർട്ട് ചെയ്യുന്നു.",
+ "Exported Pth file": "എക്സ്പോർട്ട് ചെയ്ത Pth ഫയൽ",
+ "Exported Index file": "എക്സ്പോർട്ട് ചെയ്ത ഇൻഡെക്സ് ഫയൽ",
+ "Select the pth file to be exported": "എക്സ്പോർട്ട് ചെയ്യാൻ ആദ്യം pth ഫയൽ തിരഞ്ഞെടുക്കുക",
+ "Select the index file to be exported": "എക്സ്പോർട്ട് ചെയ്യാൻ ആദ്യം ഇൻഡെക്സ് ഫയൽ തിരഞ്ഞെടുക്കുക",
+ "Upload": "അപ്ലോഡ്",
+
+ "Voice Model": "വോയ്സ് മോഡൽ",
+ "Select the voice model to use for the conversion.": "കണ്വേർഷനിനായി ഉപയോഗിക്കാൻ വോയ്സ് മോഡലുകൾ തിരഞ്ഞെടുക്കുക.",
+ "Index File": "ഇൻഡെക്സ് ഫയൽ",
+ "Select the index file to use for the conversion.": "കണ്വേർഷനിനായി ഉപയോഗിക്കേണ്ട ഇൻഡെക്സ് ഫയലുകൾ തിരഞ്ഞെടുക്കുക.",
+ "Refresh": "പുനഃസ്വന്തമാക്കുക",
+ "Unload Voice": "വോയ്സ് അൺലോഡ്",
+ "Single": "ഏകത്വം",
+ "Upload Audio": "ഓഡിയോ അപ്ലോഡ് ചെയ്യുക",
+ "Select Audio": "ഓഡിയോ തിരഞ്ഞെടുക്കുക",
+ "Select the audio to convert.": "കണ്വേർട്ട് ചെയ്യാൻ ഓഡിയോ തിരഞ്ഞെടുക്കുക.",
+ "Advanced Settings": "പുതുമയായ സെറ്റിംഗുകൾ",
+ "Clear Outputs (Deletes all audios in assets/audios)": "പരിമാറ്റുക (assets/audios എല്ലാ ഓഡിയോകൾ ഇല്ലാതാക്കുക)",
+ "Custom Output Path": "കസ്റ്റം ഔട്ട്പുട്ട് പാത",
+ "Output Path": "ഔട്ട്പുട്ട് പാത",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "ഓട്ട്പുട്ട് ഓഡിയോ സേവ്ചെയ്യപ്പെടുന്നത്, സ്വഭാവമായി assets/audios/output.wav ഇല്",
+ "Split Audio": "ഓഡിയോ വിഭാഗീകരണം",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "അനുമാനത്തിന് കൂടുതൽ ഫലങ്ങൾ ലഭിക്കാൻ ഓഡിയോ ഭാഗങ്ങൾക്ക് വിഭാഗീകരണം ചെയ്യുക.",
+ "Autotune": "ഓട്ടോട്യൂൺ",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "സോഫ്റ്റ് ഓട്ടോട്യൂൺ ആപ്ലയുകയും, സിംഗിങ് കൺവേർഷനുകളില് ശിഫാരസ് ചെയ്യുകയും.",
+ "Clean Audio": "ശുദ്ധമായ ഓഡിയോ",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "ശുദ്ധമായി നോയിസ് ഡിറ്റക്ഷൻ ആൾഗോരിതങ്ങൾ ഉപയോഗിച്ച് നിനക്ക് എത്ര പ്രയോജനപ്രദമായ ഓഡിയോ പരിഷ്കരിക്കാൻ, സ്പീക്കിംഗ് ഓഡിയോക്കിന് ശിഫാരസ് ചെയ്യുക.",
+ "Clean Strength": "ശുദ്ധി ശക്തി",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "നിനക്ക് അവശ്യമായ ഓഡിയോക്ക് ശുദ്ധിയുടെ നില സജ്ജീവമാക്കുക, അതെക്കും കൂടുതൽ ഉള്ളതും അതിനെക്കുറിച്ച് ചോദിക്കുന്നതെന്തെങ്കിലും ശുദ്ധി ചെയ്തിരിക്കുന്ന ഓഡിയോ കമ്പ്രസ്ഡ് ആയിരിക്കാനുള്ള സാധ്യതയുണ്ട്.",
+ "Pitch": "പിച്ച്",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ഓഡിയോയുടെ പിച്ച് സജ്ജീവമാക്കുക, അതെക്കും ഉയരുന്നുവെങ്കിലും പിച്ച്.",
+ "Filter Radius": "ഫിൽട്ടർ റേഡിയസ്",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "സംഖ്യ 3 അല്ലെങ്കിൽ അതിനേക്കാൾ കൂടുതൽ ആയിരിക്കുന്നാല്, ശ്വസനം കുറയ്ക്കാന് ശേഷിക്കുന്ന രീതിയില് കൂടുതല് കഴിവുള്ളതാണ്.",
+ "Search Feature Ratio": "സേര്ച്ച് ഫീച്ചർ റേഷ്യോ",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "ഇനഡെക്സ് ഫയലായി വികസിക്കുന്ന പ്രഭാവം; ഉയര്ന്ന മൂല്യം ഉയരത്തിന് അനുബന്ധ പ്രഭാവമുള്ളതാണ്. എനിക്ക് കുറഞ്ഞ മൂല്യങ്ങളെ അനുവദിക്കാനും ആര്ടിഫാക്ടുകള് നിലവിലുള്ള ഓഡിയോയിലെ ബുദ്ധിമുട്ടുകൾ ഉപയോഗപ്പെടുന്നു.",
+ "Volume Envelope": "വോള്യൂം എൻവലപ്പ്",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "ആഉട്ട്പുട്ട് ഒറ്റവന്നേറ്റത്തിന്റെ വോള്യൂം എൻവലപ്പ് സ്ഥലപ്പെടുത്തുക. അനുഭവം 1-ക്കു സമീപമായിരിക്കുന്നതും, അനുഭവ എൻവലപ്പ് ഉപയോഗപ്പെടുന്നതും കൂടുതൽ ആണ്.",
+ "Protect Voiceless Consonants": "വോയ്സ്ലസ് കോൺസനന്റുകൾ സംരക്ഷിക്കുക",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "എല്ലാവര്ക്കും പ്രകടമായ കോൺസനന്റുകൾ ഒഴുകുന്നത് എന്നതുകൊണ്ടുതന്നെ ഇലക്ട്രോ-ഓക്കുസ്റ്റിക് കൊതിയും മറ്റു ആർട്ടിഫാക്ടുകളും പ്രതിരക്ഷിക്കുന്നതിനുള്ള അരികോട്. പാരാമീറ്ററിനെ അതിന്റെ 0.5 എന്നേക്കും ഉച്ചക്കൊണ്ട് കൂട്ടിക്കൊള്ളൽ സാമൂഹ്യപ്രതിരക്ഷ നൽകുന്നു. എന്നാല്, ഈ മൂല്യം കുറഞ്ഞാക്കാൻ സാധ്യതയുണ്ട്, പ്രതിരക്ഷണം താഴെ കുറഞ്ഞുകൂടാൻ സഹായകരമാവുക.",
+ "Pitch extraction algorithm": "പിച്ച് എക്സ്ട്രാക്ഷൻ ആൾഗോരിതം",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ഓഡിയോ കൺവേർഷനിനായി ഉപയോഗിക്കേണ്ട പിച്ച് എക്സ്ട്രാക്ഷൻ ആൾഗോരിതം. സ്വതന്ത്ര ആൾഗോരിതത്താണ് rmvpe, അത് പലതരത്തിലുള്ള പ്രസ്താവനകളില് ശിഫാരസ് ചെയ്യപ്പെടുന്നു.",
+
+ "Convert": "കൺവേർട്ട് ചെയ്യുക",
+ "Export Audio": "ഓഡിയോ എക്സ്പോർട്ട് ചെയ്യുക",
+
+ "Batch": "ബാച്ച്",
+ "Input Folder": "ഇൻപുട്ട് ഫോൾഡർ",
+ "Select the folder containing the audios to convert.": "കൺവേർട്ട് ചെയ്യാൻ ഓഡിയോകളെ കാണുന്ന ഫോൾഡർ തിരഞ്ഞെടുക്കുക.",
+ "Enter input path": "ഇൻപുട്ട് പാത നൽകുക",
+ "Output Folder": "ഔട്ട്പുട്ട് ഫോൾഡർ",
+ "Select the folder where the output audios will be saved.": "ഔട്ട്പുട്ട് ഓഡിയോകൾ സേവ്ചെയ്യപ്പെടുന്ന ഫോൾഡർ തിരഞ്ഞെടുക്കുക.",
+ "Enter output path": "ഔട്ട്പുട്ട് പാത നൽകുക",
+
+ "Get information about the audio": "ഓഡിയോയുടെ കുറിപ്പ് നേടുക",
+
+ "## Voice Blender": "## വോയ്സ് ബ്ലെന്ഡർ",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "രണ്ട് വോയ്സ് മോഡലുകൾ തിരഞ്ഞെടുക്കുക, നിനക്ക് ആഗ്രഹിക്കുന്ന ബ്ലെന്റ് ശതകം സജ്ജീവമാക്കുക, അവയുടെ ബ്ലെന്റും പൂർണ്ണമായും പുതിയ ഒരു വോയ്സായാക്കുക.",
+ "Voice Blender": "വോയ്സ് ബ്ലെന്ഡർ",
+ "Drag and drop your model here": "നിനക്ക് ശൈലിക്കുകയോരോ മോഡൽ ഇവിടെ വികസിപ്പിക്കുക",
+ "You can also use a custom path.": "നിനക്ക് ഒരു സ്വന്തമായ പാതയും ഉപയോഗിക്കാം.",
+ "Blend Ratio": "ബ്ലെന്റ് അനുപാതം",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "സ്ഥാനം കൊണ്ടുകൂടുതൽ പ്രതിരൂപമാക്കാൻ മുന്നേറ്റം ഒന്നിലേറ്റെത്തിനു അല്ലെങ്കിൽ മറ്റൊന്നിലേറ്റെത്തിനു സാധിക്കും.",
+ "Fusion": "ഫ്യൂഷൻ",
+ "Path to Model": "മോഡലിന്റെ പാത",
+ "Enter path to model": "മോഡലിനെ സജ്ജീവമാക്കാൻ പാത നൽകുക",
+ "Model information to be placed": "പ്ലേസ്മെന്റ് ചെയ്യാൻ ആവശ്യമായ മോഡലിന്റെ വിവരം",
+ "Introduce the model information": "മോഡലിന്റെ വിവരം പരിചയപ്പെടുക",
+ "The information to be placed in the model (You can leave it blank or put anything).": "മോഡലില് സ്ഥലപ്പെടുത്താനുള്ള വിവരം (നിനക്ക് വിടാവുകയും അല്ലെങ്കിൽ എന്തെങ്കിലും ചേരുകയും ചെയ്യാം).",
+ "View model information": "മോഡലിന്റെ വിവരം കാണുക",
+ "Introduce the model pth path": "മോഡലിന്റെ pth പാത പരിചയപ്പെടുക",
+ "View": "കാണുക",
+ "Model extraction": "മോഡൽ എക്സ്ട്രാക്ഷൻ",
+ "Model conversion": "മോഡൽ കൺവേർഷൻ",
+ "Pth file": "Pth ഫയൽ",
+ "Output of the pth file": "Pth ഫയലിന്റെ പ്രോഡക്റ്റ്",
+ "# How to Report an Issue on GitHub": "# GitHub-ലെ ഒരു ഇഷ്യൂ റിപ്പോർട്ട് ചെയ്യുന്നതിനുള്ള രീതി",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. നിന്റെ അനുഭവപ്പെടുന്ന ഇഷ്യൂ റെക്കോർഡുചെയ്യുന്നതിന് താഴെ 'Record Screen' ബട്ടൺ ക്ലിക്കുചെയ്യുക.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. ഇഷ്യൂ റെക്കോർഡുചെയ്തുതീർക്കുന്നതിനുശേഷം, 'Stop Recording' ബട്ടൺ ക്ലിക്കുചെയ്യുക (അത് തുടർന്നിരിക്കുന്നുമോ എന്ന് താഴെയോ കാണുന്ന ലേബല് അനുസരിച്ച് മാറുന്നു).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub Issues](https://github.com/IAHispano/Applio/issues) സ്ഥലത്തേക്ക് പോകുക, 'New Issue' ബട്ടൺ ക്ലിക്കുചെയ്യുക.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. നൽകിയ ഇഷ്യൂ ടെംപ്ലേറ്റ് പൂർത്തിയാക്കുക, ആവശ്യമായ വിവരങ്ങളെ ചേർക്കുന്നതിനുശേഷം, പഴയ ഘടനയിൽ റെക്കോർഡുചെയ്ത ഫയൽ അപ്ലോഡ് ചെയ്യുന്നതിന് എസെറ്റ് വിഭാഗം ഉപയോഗിക്കുക.",
+ "Record Screen": "റെക്കോർഡ് സ്ക്രീൻ",
+ "Record": "റെക്കോർഡ്",
+ "Stop Recording": "റെക്കോർഡുനിർത്തുക",
+ "Introduce the model .pth path": "മോഡൽ .pth പാത പരിചയപ്പെടുക",
+ "See Model Information": "മോഡൽ വിവരങ്ങൾ കാണുക",
+ "## Download Model": "## മോഡൽ ഡൗൺലോഡ്",
+ "Model Link": "മോഡൽ ലിങ്ക്",
+ "Introduce the model link": "മോഡൽ ലിങ്ക് പരിചയപ്പെടുക",
+ "Download Model": "മോഡൽ ഡൗൺലോഡ്",
+ "## Drop files": "## ഫയലുകൾ ഇടുക",
+ "## Search Model": "## മോഡൽ തിരയൽ",
+ "Search": "തിരയൽ",
+ "Introduce the model name to search.": "തിരയുന്നതിനായി മോഡൽ പേര് അറിയിക്കുക.",
+ "We couldn't find models by that name.": "അനുബന്ധമായ പേരിൽ മോഡൽസ് കണ്ടെത്താനായില്ല.",
+
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "നിനക്ക് .pth ഫയലുകളും .index ഫയലുകളും ഇവിടെ ഡ്രാഗ് ചെയ്യുക. ഒന്നുകിട്ട് പിന്നെ മറ്റൊന്നു ഡ്രാഗ് ചെയ്യുക.",
+ "TTS Voices": "TTS വോയ്സുകൾ",
+ "Select the TTS voice to use for the conversion.": "മാറ്റത്തിനായി ഉപയോഗിക്കാൻ TTS വോയ്സ് തിരഞ്ഞെടുക്കുക.",
+ "Text to Synthesize": "സിന്തിയസൈസ് ചെയ്യുന്ന ടെക്സ്റ്റ്",
+ "Enter the text to synthesize.": "സിന്തിയസൈസ് ചെയ്യാൻ ടെക്സ്റ്റ് നൽകുക.",
+ "Or you can upload a .txt file": "അല്ലെങ്കിൽ .txt ഫയൽ അപ്ലോഡ് ചെയ്യാം",
+ "Enter text to synthesize": "സിന്തിയസൈസ് ചെയ്യാൻ ടെക്സ്റ്റ് നൽകുക",
+ "Output Path for TTS Audio": "TTS ഓഡിയോക്ക് ഔട്ട്പുട്ട് പാത",
+ "Output Path for RVC Audio": "RVC ഓഡിയോക്ക് ഔട്ട്പുട്ട് പാത",
+ "Enable Applio integration with Discord presence": "Discord പ്രസന്നതയോട് Applio ഇൻറഗ്രേഷൻ സജീവമാക്കുക",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "ഇത് Discord-നായിരിക്കുന്ന നിലാവ് കാണാനുള്ള സാധ്യത സജീവമാക്കും.",
+ "Enable Applio integration with applio.org/models using flask": "flask ഉപയോഗിച്ച് applio.org/models ഇൻറഗ്രേഷൻ Applio സജീവമാക്കുക",
+ "It will activate the possibility of downloading models with a click from the website.": "ഇത് വെബ്സൈറ്റിൽ ഒരു ക്ലിക്ക് ചെയ്യുമ്പോൾ മോഡലുകൾ ഡൗൺലോഡ് ചെയ്യാനുള്ള സാധ്യത സജീവമാക്കും.",
+ "Enable fake GPU": "വഞ്ചി ജിപിയു ഇയക്കുക",
+ "Training is currently unsupported due to the absence of a GPU. To activate the training tab, navigate to the settings tab and enable the 'Fake GPU' option.": "പ്രശിക്ഷണം തറന്ന് നിലവിലുള്ളതിന് ജിപിയു ഇല്ലാതെ പ്രസ്താവിക്കുന്നതിനായി തിരഞ്ഞെടുക്കുന്നത് അനുവദിക്കാൻ 'ഫെയ്ക് ജിപിയു' ഓപ്ഷൻ സജ്ജമാക്കുക എന്ന് ക്രമീകരിക്കാൻ തിരിച്ചുവരുക. ",
+ "Activates the train tab. However, please note that this device lacks GPU capabilities, hence training is not supported. This option is only for testing purposes. (This option will restart Applio)": "പ്രശിക്ഷണം തുടങ്ങുന്നു. എങ്കിലും, ദയവായി ശ്രദ്ധിക്കുക എന്നത് നിങ്ങളുടെ ഉപകരണത്തിൽ GPU സാധ്യതകൾ ഇല്ലാത്തതാണ്, അതിനാൽ പ്രശിക്ഷണം അനുവദനീയമല്ല. ഈ ഓപ്ഷൻ ഇപ്പോൾ പരീക്ഷണങ്ങളിക്കായാണ്. (ഈ ഓപ്ഷൻ അപ്ലിയോ പുനഃസജ്ജമാക്കും)",
+ "Theme": "തീം",
+ "Select the theme you want to use. (Requires restarting Applio)": "നിനക്ക് ഉപയോഗിക്കാൻ എന്താണെന്ന് നിങ്ങളുടെ തീം തിരഞ്ഞെടുക്കുക. (Applio പുനഃസജീവമാക്കാൻ ആവശ്യമാണ്)",
+ "Language": "ഭാഷ",
+ "Select the language you want to use. (Requires restarting Applio)": "നിങ്ങളുടെ ഉപയോഗത്തിന് നിങ്ങളുടെ ഭാഷ തിരഞ്ഞെടുക്കുക. (Applio പുനഃസജീവമാക്കാൻ ആവശ്യമാണ്)",
+ "Plugin Installer": "പ്ലഗിൻ ഇൻസ്റ്റാൾലർ",
+ "Drag your plugin.zip to install it": "അതിനായി നിനക്ക് നിന്നെത്തിയ .zip ഫയല് ഇൻസ്റ്റാൾ ചെയ്യാൻ ഇവിടെ ഡ്രാഗ് ചെയ്യുക",
+ "Version Checker": "വേര്ഷന് ചെക്കർ",
+ "Check which version of Applio is the latest to see if you need to update.": "നിനക്ക് അപ്ഡേറ്റുചെയ്യേണ്ടതോ എന്ന് അപ്ലിയോയുടെ ഏറ്റവും പുതിയ പതിപ്പായത് പരിശോധിക്കുക.",
+ "Check for updates": "അപ്ഡേറ്റുകൾ പരിശോധിക്കുക"
+}
diff --git a/assets/i18n/languages/mr_MR.json b/assets/i18n/languages/mr_MR.json
new file mode 100644
index 0000000000000000000000000000000000000000..012b0b6dc50f281dc89ac2d90594beacc072d090
--- /dev/null
+++ b/assets/i18n/languages/mr_MR.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "अल्टिमेट व्हॉइस क्लोनिंग टूल, अप्रतिम शक्ती, मॉड्युलरिटी आणि वापरकर्ता-अनुकूल अनुभवासाठी काळजीपूर्वक ऑप्टिमाइझ केलेले.",
+ "This section contains some extra utilities that often may be in experimental phases.": "या विभागात काही अतिरिक्त उपयोगिता आहेत ज्या बर्याचदा प्रायोगिक टप्प्यात असू शकतात.",
+ "Output Information": "आउटपुट माहिती",
+ "The output information will be displayed here.": "आउटपुट माहिती येथे प्रदर्शित केली जाईल.",
+ "Inference": "निष्कर्ष",
+ "Train": "आगगाडी",
+ "Extra": "अतिरिक्त",
+ "Merge Audios": "Merges Tऑडिओ",
+ "Processing": "प्रोसेसिंग",
+ "Audio Analyzer": "ऑडिओ विश्लेषक",
+ "Model Information": "मॉडेल माहिती",
+ "Plugins": "प्लगइन्स",
+ "Download": "डाउनलोड करा",
+ "Report a Bug": "बग ची नोंद करा",
+ "Settings": "सेटिंग्स",
+ "Preprocess": "पूर्वप्रक्रिया",
+ "Model Name": "मॉडेलचे नाव",
+ "Name of the new model.": "नवीन मॉडेलचे नाव .",
+ "Enter model name": "मॉडेल नाव प्रविष्ट करा",
+ "Dataset Path": "डेटासेट पथ",
+ "Path to the dataset folder.": "डेटासेट फोल्डरचा मार्ग.",
+ "Refresh Datasets": "डेटासेट रिफ्रेश करा",
+ "Dataset Creator": "डेटासेट निर्माता",
+ "Dataset Name": "डेटासेट चे नाव",
+ "Name of the new dataset.": "नवीन डेटासेटचे नाव.",
+ "Enter dataset name": "डेटासेट नाव प्रविष्ट करा",
+ "Upload Audio Dataset": "ऑडिओ डेटासेट अपलोड करा",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ऑडिओ फाइल डेटासेटमध्ये यशस्वीरित्या जोडली गेली आहे. कृपया प्रीप्रोसेस बटणावर क्लिक करा.",
+ "Enter dataset path": "डेटासेट मार्ग प्रविष्ट करा",
+ "Sampling Rate": "नमुना दर",
+ "The sampling rate of the audio files.": "ऑडिओ फायलींचे नमुने घेण्याचा दर.",
+ "RVC Version": "आरव्हीसी आवृत्ती",
+ "The RVC version of the model.": "मॉडेलची आरव्हीसी आवृत्ती.",
+ "Preprocess Dataset": "Preprocess Dataset",
+ "Extract": "अर्क",
+ "Hop Length": "हॉप लांबी",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "प्रणालीला महत्त्वपूर्ण खेळपट्टी बदलाकडे संक्रमण होण्यासाठी लागणारा कालावधी दर्शवितो. लहान हॉप लांबीसाठी अनुमानासाठी अधिक वेळ लागतो परंतु खेळपट्टीची अचूकता जास्त असते.",
+ "Batch Size": "बॅच आकार",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "हे आपल्या जीपीयूच्या उपलब्ध व्हीआरएएमसह संरेखित करण्याचा सल्ला दिला जातो. 4 ची सेटिंग सुधारित अचूकता परंतु हळू प्रक्रिया प्रदान करते, तर 8 वेगवान आणि मानक परिणाम प्रदान करते.",
+ "Save Every Epoch": "प्रत्येक युग वाचवा",
+ "Determine at how many epochs the model will saved at.": "मॉडेल किती युगात जतन करेल हे ठरवा.",
+ "Total Epoch": "एकूण युग",
+ "Specifies the overall quantity of epochs for the model training process.": "मॉडेल प्रशिक्षण प्रक्रियेसाठी युगांची एकूण मात्रा निर्दिष्ट करते.",
+ "Pretrained": "पूर्वप्रशिक्षित",
+ "Save Only Latest": "फक्त लेटेस्ट सेव्ह करा",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "ही सेटिंग सक्षम केल्याने जी आणि डी फायली केवळ त्यांच्या अलीकडील आवृत्त्या वाचवतील आणि स्टोरेज स्पेस प्रभावीपणे जतन करतील.",
+ "Save Every Weights": "प्रत्येक वजन वाचवा",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "ही सेटिंग आपल्याला प्रत्येक युगाच्या शेवटी मॉडेलची वजने वाचविण्यास सक्षम करते.",
+ "Custom Pretrained": "सानुकूल पूर्वप्रशिक्षित",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "सानुकूल पूर्वप्रशिक्षित मॉडेल्स चा वापर केल्यास उत्कृष्ट परिणाम मिळू शकतात, कारण विशिष्ट वापर प्रकरणानुसार सर्वात योग्य पूर्वप्रशिक्षित मॉडेल्स ची निवड केल्यास कार्यक्षमता लक्षणीय वाढू शकते.",
+ "Upload Pretrained Model": "पूर्वप्रशिक्षित मॉडेल अपलोड करा",
+ "Refresh Custom Pretraineds": "रिफ्रेश सानुकूल पूर्वप्रशिक्षित",
+ "Pretrained Custom Settings": "पूर्वप्रशिक्षित सानुकूल सेटिंग्ज",
+ "The file you dropped is not a valid pretrained file. Please try again.": "आपण टाकलेली फाईल वैध पूर्वप्रशिक्षित फाइल नाही. कृपया पुन्हा प्रयत्न करा.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "ड्रॉपडाउन मेनूमध्ये पूर्वप्रशिक्षित फाइल पाहण्यासाठी रिफ्रेश बटणावर क्लिक करा.",
+ "Pretrained G Path": "सानुकूल पूर्वप्रशिक्षित जी",
+ "Pretrained D Path": "सानुकूल पूर्वप्रशिक्षित डी",
+ "GPU Settings": "जीपीयू सेटिंग्स",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "प्रगत जीपीयू सेटिंग्ज सेट करा, चांगल्या जीपीयू आर्किटेक्चर असलेल्या वापरकर्त्यांसाठी शिफारस केली आहे.",
+ "GPU Custom Settings": "जीपीयू सानुकूल सेटिंग्स",
+ "GPU Number": "जीपीयू नंबर",
+ "0 to ∞ separated by -": "0 ते ∞ वेगळे केले जातात -",
+ "GPU Information": "जीपीयू माहिती",
+ "Pitch Guidance": "खेळपट्टी मार्गदर्शन",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "खेळपट्टीमार्गदर्शनाचा वापर करून मूळ आवाजाच्या खेळपट्टीसह त्याच्या अंतर्मुखतेचे प्रतिबिंब उमटविणे शक्य होते. हे वैशिष्ट्य गायन आणि इतर दृश्यांसाठी विशेषतः मौल्यवान आहे जिथे मूळ सूर किंवा पिच पॅटर्न जतन करणे आवश्यक आहे.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "स्वत: चे प्रशिक्षण देताना पूर्वप्रशिक्षित मॉडेल्स वापरा. हा दृष्टिकोन प्रशिक्षण कालावधी कमी करतो आणि एकंदर गुणवत्ता वाढवतो.",
+ "Extract Features": "अर्क वैशिष्ट्ये",
+ "Start Training": "प्रशिक्षण सुरू करा",
+ "Generate Index": "इंडेक्स तयार करा",
+ "Voice Model": "व्हॉइस मॉडेल",
+ "Select the voice model to use for the conversion.": "रूपांतरणासाठी वापरण्यासाठी व्हॉइस मॉडेल निवडा.",
+ "Index File": "अनुक्रमणिका फाइल",
+ "Select the index file to use for the conversion.": "रूपांतरणासाठी वापरण्यासाठी अनुक्रमणिका फाइल निवडा.",
+ "Refresh": "टवटवी आणणे",
+ "Unload Voice": "आवाज अनलोड करा",
+ "Single": "अविवाहित",
+ "Upload Audio": "ऑडिओ अपलोड करा",
+ "Select Audio": "ऑडिओ निवडा",
+ "Select the audio to convert.": "रूपांतरित करण्यासाठी ऑडिओ निवडा.",
+ "Advanced Settings": "प्रगत सेटिंग्ज",
+ "Clear Outputs (Deletes all audios in assets/audios)": "स्पष्ट आउटपुट (मालमत्ता / ऑडिओमधील सर्व ऑडिओ हटवतात)",
+ "Custom Output Path": "सानुकूल आउटपुट पथ",
+ "Output Path": "आउटपुट पथ",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "तो मार्ग जिथे आउटपुट ऑडिओ जतन केला जाईल, मालमत्ता / ऑडिओ / output.wav मध्ये डिफॉल्टद्वारे",
+ "Split Audio": "स्प्लिट ऑडिओ",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "काही प्रकरणांमध्ये चांगले परिणाम मिळविण्यासाठी अनुमानासाठी ऑडिओचे तुकडे करा.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "गायन रूपांतरणासाठी शिफारस केलेल्या आपल्या निष्कर्षांवर सॉफ्ट ऑटोट्यून लागू करा.",
+ "Clean Audio": "स्वच्छ ऑडिओ",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "ध्वनी शोध अल्गोरिदम वापरुन आपले ऑडिओ आउटपुट स्वच्छ करा, ऑडिओ बोलण्यासाठी शिफारस केली आहे.",
+ "Clean Strength": "स्वच्छ शक्ती",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "आपल्याला हव्या असलेल्या ऑडिओमध्ये क्लीन-अप पातळी सेट करा, आपण ते जितके वाढवाल तितके ते साफ होईल, परंतु ऑडिओ अधिक संकुचित होण्याची शक्यता आहे.",
+ "Pitch": "खेळपट्टी",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ऑडिओची खेळपट्टी सेट करा, किंमत जितकी जास्त तितकी खेळपट्टी जास्त.",
+ "Filter Radius": "फिल्टर त्रिज्या",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "जर संख्या तीनपेक्षा जास्त किंवा समान असेल तर संकलित टोन परिणामांवर मध्यम फिल्टरिंग वापरल्यास श्वसन कमी होण्याची क्षमता असते.",
+ "Search Feature Ratio": "शोध वैशिष्ट्य गुणोत्तर",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "अनुक्रमणिका संचिकेने टाकलेला प्रभाव; उच्च मूल्य अधिक प्रभावाशी संबंधित आहे. तथापि, कमी मूल्यांची निवड केल्याने ऑडिओमध्ये असलेल्या कलाकृती कमी होण्यास मदत होते.",
+ "Volume Envelope": "Volume लिफाफा",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "आउटपुटच्या व्हॉल्यूम लिफाफ्यासह पर्याय किंवा मिश्रण करा. गुणोत्तर 1 च्या जितके जवळ असेल तितके आउटपुट लिफाफा वापरला जातो.",
+ "Protect Voiceless Consonants": "आवाजहीन व्यंजनांचे रक्षण करा",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "इलेक्ट्रो-ध्वनिक फाटणे आणि इतर कलाकृती टाळण्यासाठी विशिष्ट व्यंजने आणि श्वासोच्छवासाच्या आवाजांचे रक्षण करा. पॅरामीटरला त्याच्या 0.5 च्या जास्तीत जास्त मूल्यावर खेचणे व्यापक संरक्षण प्रदान करते. तथापि, हे मूल्य कमी केल्याने अनुक्रमणिका प्रभाव संभाव्यत: कमी करताना संरक्षणाची व्याप्ती कमी होऊ शकते.",
+ "Pitch extraction algorithm": "पिच निष्कर्षण अल्गोरिदम",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ऑडिओ रूपांतरणासाठी वापरण्यासाठी पिच एक्सट्रॅक्शन अल्गोरिदम. डिफॉल्ट अल्गोरिदम आरएमव्हीपीई आहे, ज्याची शिफारस बहुतेक प्रकरणांमध्ये केली जाते.",
+ "Convert": "धर्मांतर करा",
+ "Export Audio": "निर्यात ऑडिओ",
+ "Batch": "तुकडी",
+ "Input Folder": "इनपुट फोल्डर",
+ "Select the folder containing the audios to convert.": "रूपांतरित करण्यासाठी ऑडिओ असलेले फोल्डर निवडा.",
+ "Enter input path": "इनपुट पथ प्रविष्ट करा",
+ "Output Folder": "आउटपुट फोल्डर",
+ "Select the folder where the output audios will be saved.": "फोल्डर निवडा जेथे आउटपुट ऑडिओ सेव्ह केले जातील.",
+ "Enter output path": "आउटपुट पथ प्रविष्ट करा",
+ "Get information about the audio": "ऑडिओबद्दल माहिती मिळवा",
+ "Information about the audio file": "ऑडिओ फाईलची माहिती",
+ "Waiting for information...": "माहितीच्या प्रतीक्षेत...",
+ "## Voice Blender": "## व्हॉइस ब्लेंडर",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "दोन व्हॉईस मॉडेल निवडा, आपली इच्छित मिश्रण टक्केवारी सेट करा आणि त्यांना पूर्णपणे नवीन आवाजात मिसळा.",
+ "Voice Blender": "व्हॉइस ब्लेंडर",
+ "Drag and drop your model here": "आपले मॉडेल येथे खेचून टाका आणि सोडा",
+ "You can also use a custom path.": "आपण सानुकूल मार्ग देखील वापरू शकता.",
+ "Blend Ratio": "मिश्रण गुणोत्तर",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "एका किंवा दुसर्या बाजूला स्थिती अधिक समायोजित केल्याने मॉडेल पहिल्या किंवा दुसर्या बाजूला अधिक समान होईल.",
+ "Fusion": "फ्यूजन",
+ "Path to Model": "मॉडेल चा मार्ग",
+ "Enter path to model": "मॉडेलचा मार्ग प्रविष्ट करा",
+ "Model information to be placed": "मॉडेल माहिती ठेवली जाईल",
+ "Inroduce the model information": "मॉडेल ची माहिती सादर करा",
+ "The information to be placed in the model (You can leave it blank or put anything).": "मॉडेलमध्ये ठेवली जाणारी माहिती (आपण ती रिकामी ठेवू शकता किंवा काहीही टाकू शकता).",
+ "View model information": "मॉडेल माहिती पहा",
+ "Introduce the model pth path": "मॉडेल पीटीएच पथ सादर करा",
+ "View": "पहा",
+ "Model extraction": "मॉडेल निष्कर्षण",
+ "Model conversion": "मॉडेल रूपांतरण",
+ "Pth file": "पीटीएच फाईल",
+ "Output of the pth file": "पीटीएच फाईलचे आउटपुट",
+ "# How to Report an Issue on GitHub": "# गिटहबवर एखाद्या समस्येची नोंद कशी करावी",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. आपण अनुभवत असलेल्या समस्येचे रेकॉर्डिंग सुरू करण्यासाठी खालील 'रेकॉर्ड स्क्रीन' बटणावर क्लिक करा.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "२. अंक रेकॉर्ड िंग पूर्ण झाल्यावर 'स्टॉप रेकॉर्डिंग' बटणावर क्लिक करा (तेच बटण, पण तुम्ही सक्रियपणे रेकॉर्डिंग करत आहात की नाही यावर अवलंबून लेबल बदलते).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [गिटहब इश्यूज] (https://github.com/IAHispano/Applio/issues) वर जा आणि 'न्यू इश्यू' बटणावर क्लिक करा.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. प्रदान केलेले इश्यू टेम्पलेट पूर्ण करा, आवश्यकतेनुसार तपशील समाविष्ट करण्याची खात्री करा आणि मागील चरणातून रेकॉर्ड केलेली फाइल अपलोड करण्यासाठी मालमत्ता विभागाचा वापर करा.",
+ "Record Screen": "रेकॉर्ड स्क्रीन",
+ "Record": "नोंदणे",
+ "Stop Recording": "रेकॉर्डिंग थांबवा",
+ "Introduce the model .pth path": "मॉडेल .पीटीएच पथ सादर करा",
+ "See Model Information": "मॉडेल माहिती पहा",
+ "## Download Model": "## मॉडेल डाऊनलोड करा",
+ "Model Link": "मॉडेल लिंक",
+ "Introduce the model link": "मॉडेल लिंक ची ओळख करून द्या",
+ "Download Model": "मॉडेल डाउनलोड करा",
+ "## Drop files": "## फाइल्स ड्रॉप करा",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "आपली .pth फाइल आणि .अनुक्रमणिका फाईल या जागेत खेचून घ्या. एकाला ओढून घ्या आणि मग दुसरं.",
+ "TTS Voices": "टीटीएस आवाज",
+ "Select the TTS voice to use for the conversion.": "रूपांतरणासाठी वापरण्यासाठी टीटीएस आवाज निवडा.",
+ "Text to Synthesize": "संश्लेषित करण्यासाठी मजकूर",
+ "Enter the text to synthesize.": "संश्लेषित करण्यासाठी मजकूर प्रविष्ट करा.",
+ "Or you can upload a .txt file": "किंवा तुम्ही .txt फाईल अपलोड करू शकता",
+ "Enter text to synthesize": "संश्लेषित करण्यासाठी मजकूर प्रविष्ट करा",
+ "Output Path for TTS Audio": "टीटीएस ऑडिओसाठी आउटपुट पथ",
+ "Output Path for RVC Audio": "आरव्हीसी ऑडिओसाठी आउटपुट पथ",
+ "Enable Applio integration with Discord presence": "डिस्कॉर्ड उपस्थितीसह अनुप्रयोग एकीकरण सक्षम करा",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "हे डिस्कॉर्डमध्ये सध्याची ऍप्लिओ क्रियाकलाप प्रदर्शित करण्याची शक्यता सक्रिय करेल.",
+ "Enable Applio integration with applio.org/models using flask": "फ्लास्क वापरुन applio.org/models अनुप्रयोगा एकीकरण सक्षम करा",
+ "It will activate the possibility of downloading models with a click from the website.": "हे वेबसाइटवरून एका क्लिकवर मॉडेल्स डाउनलोड करण्याची शक्यता सक्रिय करेल.",
+ "Theme": "थीम :",
+ "Select the theme you want to use. (Requires restarting Applio)": "आपण वापरू इच्छित थीम निवडा. (अनुप्रयोग पुन्हा सुरू करणे आवश्यक आहे)",
+ "Language": "भाषा",
+ "Select the language you want to use. (Requires restarting Applio)": "आपल्याला जी भाषा वापरायची आहे ती निवडा. (अनुप्रयोग पुन्हा सुरू करणे आवश्यक आहे)",
+ "Plugin Installer": "Plugin Installer",
+ "Drag your plugin.zip to install it": "ते स्थापित करण्यासाठी आपले plugin.zip खेचून घ्या",
+ "Version Checker": "Version Checker",
+ "Check which version of Applio is the latest to see if you need to update.": "आपल्याला अद्ययावत करण्याची आवश्यकता आहे की नाही हे पाहण्यासाठी अ ॅप्लिओची कोणती आवृत्ती नवीनतम आहे ते पहा.",
+ "Check for updates": "अपडेट्ससाठी पहा"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/ms_MS.json b/assets/i18n/languages/ms_MS.json
new file mode 100644
index 0000000000000000000000000000000000000000..c122b395a424174b36dd2efb9a88db4a26c522bd
--- /dev/null
+++ b/assets/i18n/languages/ms_MS.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Alat pengklonan suara muktamad, dioptimumkan dengan teliti untuk kuasa yang tiada tandingan, modulariti, dan pengalaman mesra pengguna.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Bahagian ini mengandungi beberapa utiliti tambahan yang selalunya berada dalam fasa percubaan.",
+ "Output Information": "Maklumat Output",
+ "The output information will be displayed here.": "Maklumat output akan dipaparkan di sini.",
+ "Inference": "Inferens",
+ "Train": "Kereta api",
+ "Extra": "Tambahan",
+ "Merge Audios": "Mencantumkan Audio",
+ "Processing": "Pemprosesan",
+ "Audio Analyzer": "Penganalisis Audio",
+ "Model Information": "Maklumat Model",
+ "Plugins": "Plugin",
+ "Download": "Muat turun",
+ "Report a Bug": "Laporkan pepijat",
+ "Settings": "Seting",
+ "Preprocess": "Praproses",
+ "Model Name": "Nama Model",
+ "Name of the new model.": "Nama model baharu.",
+ "Enter model name": "Masukkan nama model",
+ "Dataset Path": "Laluan Set Data",
+ "Path to the dataset folder.": "Laluan ke folder set data.",
+ "Refresh Datasets": "Menyegar semula Set Data",
+ "Dataset Creator": "Pencipta Set Data",
+ "Dataset Name": "Nama Set Data",
+ "Name of the new dataset.": "Nama set data baru.",
+ "Enter dataset name": "Masukkan nama set data",
+ "Upload Audio Dataset": "Muat naik Set Data Audio",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Fail audio telah berjaya ditambahkan pada set data. Sila klik butang praproses.",
+ "Enter dataset path": "Memasukkan laluan set data",
+ "Sampling Rate": "Kadar Persampelan",
+ "The sampling rate of the audio files.": "Kadar pensampelan fail audio.",
+ "RVC Version": "Versi RVC",
+ "The RVC version of the model.": "Versi RVC model.",
+ "Preprocess Dataset": "Set Data Praproses",
+ "Extract": "Cabutan",
+ "Hop Length": "Panjang Hop",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Menunjukkan tempoh yang diperlukan untuk sistem beralih kepada perubahan padang yang ketara. Panjang hop yang lebih kecil memerlukan lebih banyak masa untuk kesimpulan tetapi cenderung menghasilkan ketepatan padang yang lebih tinggi.",
+ "Batch Size": "Saiz kelompok",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Anda dinasihatkan untuk menyelaraskannya dengan VRAM GPU anda yang tersedia. Tetapan 4 menawarkan ketepatan yang lebih baik tetapi pemprosesan yang lebih perlahan, manakala 8 memberikan hasil yang lebih cepat dan standard.",
+ "Save Every Epoch": "Simpan Setiap Zaman",
+ "Determine at how many epochs the model will saved at.": "Tentukan berapa zaman model akan disimpan.",
+ "Total Epoch": "Jumlah Zaman",
+ "Specifies the overall quantity of epochs for the model training process.": "Menentukan kuantiti keseluruhan zaman untuk proses latihan model.",
+ "Pretrained": "Dipralatih",
+ "Save Only Latest": "Simpan Terkini Sahaja",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Mendayakan tetapan ini akan menyebabkan fail G dan D menyimpan hanya versi terbaru mereka, menjimatkan ruang storan dengan berkesan.",
+ "Save Every Weights": "Jimat Setiap Berat",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Tetapan ini membolehkan anda menjimatkan berat model pada akhir setiap zaman.",
+ "Custom Pretrained": "Pralatih Tersuai",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Menggunakan model pra-latihan tersuai boleh membawa kepada hasil yang unggul, kerana memilih model pra-latihan yang paling sesuai yang disesuaikan dengan kes penggunaan tertentu dapat meningkatkan prestasi dengan ketara.",
+ "Upload Pretrained Model": "Muat naik Model Pralatih",
+ "Refresh Custom Pretraineds": "Menyegar Semula Pralatihan Tersuai",
+ "Pretrained Custom Settings": "Seting Tersuai Pralatih",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Fail yang anda gugurkan bukan fail pralatih yang sah. Sila cuba lagi.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Klik butang segar semula untuk melihat fail pralatih dalam menu juntai bawah.",
+ "Pretrained G Path": "Custom Pretrained G",
+ "Pretrained D Path": "Custom Pretrained D",
+ "GPU Settings": "Seting GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Menetapkan tetapan GPU lanjutan, disyorkan untuk pengguna dengan seni bina GPU yang lebih baik.",
+ "GPU Custom Settings": "Seting Tersuai GPU",
+ "GPU Number": "Nombor GPU",
+ "0 to ∞ separated by -": "0 hingga ∞ dipisahkan oleh -",
+ "GPU Information": "Maklumat GPU",
+ "Pitch Guidance": "Panduan Padang",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Dengan menggunakan panduan padang, ia boleh dilaksanakan untuk mencerminkan intonasi suara asal, termasuk padangnya. Ciri ini amat berharga untuk nyanyian dan senario lain di mana mengekalkan corak melodi atau padang asal adalah penting.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Gunakan model yang telah dilatih semasa melatih anda sendiri. Pendekatan ini mengurangkan tempoh latihan dan meningkatkan kualiti keseluruhan.",
+ "Extract Features": "Ciri Ekstrak",
+ "Start Training": "Mulakan Latihan",
+ "Generate Index": "Menjana Indeks",
+ "Voice Model": "Model Suara",
+ "Select the voice model to use for the conversion.": "Pilih model suara untuk digunakan untuk penukaran.",
+ "Index File": "Fail Indeks",
+ "Select the index file to use for the conversion.": "Pilih fail indeks untuk digunakan untuk penukaran.",
+ "Refresh": "Refresh",
+ "Unload Voice": "Memunggah Suara",
+ "Single": "Tunggal",
+ "Upload Audio": "Muat naik Audio",
+ "Select Audio": "Pilih Audio",
+ "Select the audio to convert.": "Pilih audio untuk ditukar.",
+ "Advanced Settings": "Seting Lanjutan",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Kosongkan Output (Memadamkan semua audio dalam aset/audio)",
+ "Custom Output Path": "Laluan Output Tersuai",
+ "Output Path": "Laluan Output",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Laluan di mana audio output akan disimpan, secara lalai dalam aset/audio/output.wav",
+ "Split Audio": "Audio Pisah",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Pisahkan audio kepada ketulan untuk kesimpulan untuk mendapatkan hasil yang lebih baik dalam beberapa kes.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Gunakan autotune lembut pada kesimpulan anda, disyorkan untuk penukaran nyanyian.",
+ "Clean Audio": "Audio Bersih",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Bersihkan output audio anda menggunakan algoritma pengesanan bunyi, disyorkan untuk bercakap audio.",
+ "Clean Strength": "Kekuatan Bersih",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Tetapkan tahap pembersihan ke audio yang anda mahukan, semakin banyak anda meningkatkannya semakin banyak ia akan membersihkan, tetapi ada kemungkinan audio akan lebih dimampatkan.",
+ "Pitch": "Pitch",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Tetapkan padang audio, semakin tinggi nilai, semakin tinggi padang.",
+ "Filter Radius": "Jejari Penapis",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Sekiranya bilangannya lebih besar daripada atau sama dengan tiga, menggunakan penapisan median pada hasil nada yang dikumpulkan berpotensi untuk mengurangkan pernafasan.",
+ "Search Feature Ratio": "Nisbah Ciri Carian",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Pengaruh yang dikenakan oleh fail indeks; Nilai yang lebih tinggi sepadan dengan pengaruh yang lebih besar. Walau bagaimanapun, memilih nilai yang lebih rendah boleh membantu mengurangkan artifak yang terdapat dalam audio.",
+ "Volume Envelope": "Sampul Kelantangan",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Gantikan atau gabungkan dengan sampul kelantangan output. Semakin dekat nisbahnya kepada 1, semakin banyak sampul output digunakan.",
+ "Protect Voiceless Consonants": "Lindungi Konsonan Tanpa Suara",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Lindungi konsonan yang berbeza dan bunyi pernafasan untuk mengelakkan koyakan elektro-akustik dan artifak lain. Menarik parameter ke nilai maksimum 0.5 menawarkan perlindungan komprehensif. Walau bagaimanapun, mengurangkan nilai ini mungkin mengurangkan tahap perlindungan sambil berpotensi mengurangkan kesan pengindeksan.",
+ "Pitch extraction algorithm": "Algoritma pengekstrakan padang",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Algoritma pengekstrakan padang untuk digunakan untuk penukaran audio. Algoritma lalai ialah rmvpe, yang disyorkan untuk kebanyakan kes.",
+ "Convert": "Menukar",
+ "Export Audio": "Eksport Audio",
+ "Batch": "Kelompok",
+ "Input Folder": "Input Folder",
+ "Select the folder containing the audios to convert.": "Pilih folder yang mengandungi audio untuk ditukar.",
+ "Enter input path": "Masukkan laluan input",
+ "Output Folder": "Output Folder",
+ "Select the folder where the output audios will be saved.": "Pilih folder di mana audio output akan disimpan.",
+ "Enter output path": "Masukkan laluan output",
+ "Get information about the audio": "Mendapatkan maklumat tentang audio",
+ "Information about the audio file": "Maklumat mengenai fail audio",
+ "Waiting for information...": "Menunggu maklumat...",
+ "## Voice Blender": "## Pengisar Suara",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Pilih dua model suara, tetapkan peratusan campuran yang anda inginkan, dan gabungkannya ke dalam suara yang sama sekali baru.",
+ "Voice Blender": "Pengisar Suara",
+ "Drag and drop your model here": "Seret dan lepas model anda di sini",
+ "You can also use a custom path.": "Anda juga boleh menggunakan laluan tersuai.",
+ "Blend Ratio": "Nisbah Campuran",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Melaraskan kedudukan lebih ke arah satu sisi atau yang lain akan menjadikan model lebih serupa dengan yang pertama atau kedua.",
+ "Fusion": "Gabungan",
+ "Path to Model": "Laluan ke Model",
+ "Enter path to model": "Masukkan laluan ke model",
+ "Model information to be placed": "Maklumat model yang akan diletakkan",
+ "Inroduce the model information": "Menghasut maklumat model",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Maklumat yang akan diletakkan dalam model (Anda boleh membiarkannya kosong atau meletakkan apa-apa).",
+ "View model information": "Lihat maklumat model",
+ "Introduce the model pth path": "Perkenalkan laluan pth model",
+ "View": "Lihat",
+ "Model extraction": "Pengekstrakan model",
+ "Model conversion": "Penukaran model",
+ "Pth file": "Pth fail",
+ "Output of the pth file": "Output fail pth",
+ "# How to Report an Issue on GitHub": "# Cara Melaporkan Isu di GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Klik pada butang 'Skrin Rekod' di bawah untuk mula merakam isu yang anda alami.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Setelah anda selesai merakam isu ini, klik pada butang 'Hentikan Rakaman' (butang yang sama, tetapi label berubah bergantung pada sama ada anda merakam secara aktif atau tidak).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Pergi ke [Isu GitHub](https://github.com/IAHispano/Applio/issues) dan klik pada butang 'Isu Baru'.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Lengkapkan templat isu yang disediakan, pastikan untuk memasukkan butiran mengikut keperluan, dan gunakan bahagian aset untuk memuat naik fail yang dirakam dari langkah sebelumnya.",
+ "Record Screen": "Skrin Rakam",
+ "Record": "Rekod",
+ "Stop Recording": "Hentikan Rakaman",
+ "Introduce the model .pth path": "Memperkenalkan model laluan .pth",
+ "See Model Information": "Lihat Maklumat Model",
+ "## Download Model": "## Muat Turun Model",
+ "Model Link": "Pautan Model",
+ "Introduce the model link": "Memperkenalkan pautan model",
+ "Download Model": "Model Muat Turun",
+ "## Drop files": "## Jatuhkan fail",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Seret fail .pth dan fail .index anda ke dalam ruang ini. Seret satu dan kemudian yang lain.",
+ "TTS Voices": "Suara TTS",
+ "Select the TTS voice to use for the conversion.": "Pilih suara TTS untuk digunakan bagi penukaran.",
+ "Text to Synthesize": "Teks untuk Mensintesis",
+ "Enter the text to synthesize.": "Masukkan teks untuk mensintesiskan.",
+ "Or you can upload a .txt file": "Atau anda boleh memuat naik fail .txt",
+ "Enter text to synthesize": "Masukkan teks untuk mensintesis saiz",
+ "Output Path for TTS Audio": "Laluan output untuk TTS Audio",
+ "Output Path for RVC Audio": "Laluan Output untuk Audio RVC",
+ "Enable Applio integration with Discord presence": "Dayakan integrasi Applio dengan kehadiran Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Ia akan mengaktifkan kemungkinan memaparkan aktiviti Applio semasa di Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Dayakan integrasi Applio dengan applio.org/models menggunakan kelalang",
+ "It will activate the possibility of downloading models with a click from the website.": "Ia akan mengaktifkan kemungkinan memuat turun model dengan satu klik dari laman web.",
+ "Theme": "Tema",
+ "Select the theme you want to use. (Requires restarting Applio)": "Pilih tema yang anda mahu gunakan. (Memerlukan memulakan semula Applio)",
+ "Language": "Bahasa",
+ "Select the language you want to use. (Requires restarting Applio)": "Pilih bahasa yang anda mahu gunakan. (Memerlukan memulakan semula Applio)",
+ "Plugin Installer": "Pemasang Plugin",
+ "Drag your plugin.zip to install it": "Seret plugin.zip anda untuk memasangnya",
+ "Version Checker": "Penyemak Versi",
+ "Check which version of Applio is the latest to see if you need to update.": "Semak versi Applio yang terkini untuk melihat sama ada anda perlu mengemas kini.",
+ "Check for updates": "Semak kemas kini"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/nl_NL.json b/assets/i18n/languages/nl_NL.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fd29e132401f6611ffc02d4f44d45d4d9307a39
--- /dev/null
+++ b/assets/i18n/languages/nl_NL.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Ultieme tool voor het klonen van stemmen, zorgvuldig geoptimaliseerd voor ongeëvenaarde kracht, modulariteit en gebruiksvriendelijke ervaring.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Deze sectie bevat enkele extra hulpprogramma's die zich vaak in experimentele fasen bevinden.",
+ "Output Information": "Output Informatie",
+ "The output information will be displayed here.": "De uitvoerinformatie wordt hier weergegeven.",
+ "Inference": "Conclusie",
+ "Train": "Trein",
+ "Extra": "Extra",
+ "Merge Audios": "Audio samenvoegen",
+ "Processing": "Verwerking",
+ "Audio Analyzer": "Audio Analyzer",
+ "Model Information": "Modelinformatie",
+ "Plugins": "Insteekplaatsen",
+ "Download": "Downloaden",
+ "Report a Bug": "Een bug melden",
+ "Settings": "Instellingen",
+ "Preprocess": "Voorbewerking",
+ "Model Name": "Modelnaam",
+ "Name of the new model.": "Naam van het nieuwe model.",
+ "Enter model name": "Voer de modelnaam in",
+ "Dataset Path": "Pad naar gegevensset",
+ "Path to the dataset folder.": "Pad naar de map met gegevenssets.",
+ "Refresh Datasets": "Gegevenssets vernieuwen",
+ "Dataset Creator": "Maker van gegevenssets",
+ "Dataset Name": "Naam van gegevensset",
+ "Name of the new dataset.": "Naam van de nieuwe dataset.",
+ "Enter dataset name": "Voer de naam van de gegevensset in",
+ "Upload Audio Dataset": "Audiogegevensset uploaden",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Het audiobestand is toegevoegd aan de dataset. Klik op de knop voorbewerking.",
+ "Enter dataset path": "Pad naar gegevensset invoeren",
+ "Sampling Rate": "Bemonsteringsfrequentie",
+ "The sampling rate of the audio files.": "De bemonsteringsfrequentie van de audiobestanden.",
+ "RVC Version": "RVC-versie",
+ "The RVC version of the model.": "De RVC-versie van het model.",
+ "Preprocess Dataset": "Gegevensset voor het proces",
+ "Extract": "Extract",
+ "Hop Length": "Hop Lengte",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Geeft de tijd aan die nodig is om het systeem over te zetten op een significante toonhoogteverandering. Kleinere hoplengtes hebben meer tijd nodig om conclusies te trekken, maar hebben de neiging om een hogere toonhoogtenauwkeurigheid op te leveren.",
+ "Batch Size": "Batchgrootte",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Het is aan te raden om deze af te stemmen op het beschikbare VRAM van je GPU. Een instelling van 4 biedt verbeterde nauwkeurigheid maar langzamere verwerking, terwijl 8 snellere en standaardresultaten oplevert.",
+ "Save Every Epoch": "Red elk tijdperk",
+ "Determine at how many epochs the model will saved at.": "Bepaal op hoeveel epochs het model wordt opgeslagen.",
+ "Total Epoch": "Totale tijdvak",
+ "Specifies the overall quantity of epochs for the model training process.": "Hiermee geeft u het totale aantal epochs op voor het modeltrainingsproces.",
+ "Pretrained": "Voorgetraind",
+ "Save Only Latest": "Alleen de nieuwste opslaan",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Als u deze instelling inschakelt, worden de G- en D-bestanden alleen hun meest recente versies opgeslagen, waardoor er effectief opslagruimte wordt bespaard.",
+ "Save Every Weights": "Sla alle gewichten op",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Met deze instelling kunt u de gewichten van het model aan het einde van elk tijdperk opslaan.",
+ "Custom Pretrained": "Aangepaste voorgetrainde",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Het gebruik van op maat gemaakte voorgetrainde modellen kan tot superieure resultaten leiden, aangezien het selecteren van de meest geschikte vooraf getrainde modellen die zijn afgestemd op de specifieke gebruikssituatie de prestaties aanzienlijk kan verbeteren.",
+ "Upload Pretrained Model": "Vooraf getraind model uploaden",
+ "Refresh Custom Pretraineds": "Aangepaste vooraf getrainde bestanden vernieuwen",
+ "Pretrained Custom Settings": "Vooraf getrainde aangepaste instellingen",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Het bestand dat u hebt neergezet, is geen geldig vooraf getraind bestand. Probeer het opnieuw.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Klik op de knop Vernieuwen om het vooraf getrainde bestand in het vervolgkeuzemenu te zien.",
+ "Pretrained G Path": "Aangepaste voorgetrainde G",
+ "Pretrained D Path": "Aangepaste voorgetrainde D",
+ "GPU Settings": "GPU-instellingen",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Hiermee stelt u geavanceerde GPU-instellingen in, aanbevolen voor gebruikers met een betere GPU-architectuur.",
+ "GPU Custom Settings": "Aangepaste GPU-instellingen",
+ "GPU Number": "GPU-nummer",
+ "0 to ∞ separated by -": "0 tot ∞ gescheiden door -",
+ "GPU Information": "GPU-informatie",
+ "Pitch Guidance": "Begeleiding bij het veld",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Door gebruik te maken van toonhoogtebegeleiding wordt het mogelijk om de intonatie van de originele stem, inclusief de toonhoogte, te spiegelen. Deze functie is met name waardevol voor zang en andere scenario's waarbij het behoud van de originele melodie of het toonhoogtepatroon essentieel is.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Gebruik vooraf getrainde modellen bij het trainen van uw eigen modellen. Deze aanpak verkort de trainingsduur en verbetert de algehele kwaliteit.",
+ "Extract Features": "Extraheer functies",
+ "Start Training": "Begin met trainen",
+ "Generate Index": "Index genereren",
+ "Voice Model": "Stem Model",
+ "Select the voice model to use for the conversion.": "Selecteer het spraakmodel dat u voor de conversie wilt gebruiken.",
+ "Index File": "Index-bestand",
+ "Select the index file to use for the conversion.": "Selecteer het indexbestand dat u voor de conversie wilt gebruiken.",
+ "Refresh": "Opfrissen",
+ "Unload Voice": "Stem lossen",
+ "Single": "Ongetrouwd",
+ "Upload Audio": "Audio uploaden",
+ "Select Audio": "Selecteer Audio",
+ "Select the audio to convert.": "Selecteer de audio die u wilt converteren.",
+ "Advanced Settings": "Geavanceerde instellingen",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Uitvoer wissen (verwijdert alle audio in assets/audio)",
+ "Custom Output Path": "Aangepast uitvoerpad",
+ "Output Path": "Uitgang Pad",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Het pad waar de uitvoeraudio wordt opgeslagen, standaard in assets/audios/output.wav",
+ "Split Audio": "Audio splitsen",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Splits de audio op in stukken voor gevolgtrekking om in sommige gevallen betere resultaten te verkrijgen.",
+ "Autotune": "Automatisch afstemmen",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Pas een zachte autotune toe op je inferenties, aanbevolen voor zangconversies.",
+ "Clean Audio": "Schone audio",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Reinig uw audio-uitvoer met behulp van ruisdetectie-algoritmen, aanbevolen voor gesproken audio.",
+ "Clean Strength": "Schone kracht",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Stel het opschoningsniveau in op de gewenste audio, hoe meer u het verhoogt, hoe meer het zal opschonen, maar het is mogelijk dat de audio meer gecomprimeerd zal zijn.",
+ "Pitch": "Toonhoogte",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Stel de toonhoogte van de audio in, hoe hoger de waarde, hoe hoger de toonhoogte.",
+ "Filter Radius": "Filter Straal",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Als het getal groter is dan of gelijk is aan drie, kan het gebruik van mediane filtering op de verzamelde toonresultaten de ademhaling verminderen.",
+ "Search Feature Ratio": "Verhouding zoekfunctie",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Invloed uitgeoefend door het indexbestand; Een hogere waarde komt overeen met een grotere invloed. Als u echter voor lagere waarden kiest, kunt u de artefacten in de audio verminderen.",
+ "Volume Envelope": "Volume Envelop",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Vervang of meng met de volume-envelop van de uitvoer. Hoe dichter de verhouding bij 1 ligt, hoe meer de uitvoerenveloppe wordt gebruikt.",
+ "Protect Voiceless Consonants": "Bescherm stemloze medeklinkers",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Bescherm verschillende medeklinkers en ademhalingsgeluiden om elektro-akoestische scheuren en andere artefacten te voorkomen. Door de parameter naar de maximale waarde van 0,5 te trekken, wordt uitgebreide bescherming geboden. Het verlagen van deze waarde kan echter de mate van bescherming verminderen en mogelijk het indexeringseffect beperken.",
+ "Pitch extraction algorithm": "Algoritme voor het extraheren van toonhoogte",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Pitch-extractie-algoritme om te gebruiken voor de audioconversie. Het standaardalgoritme is rmvpe, wat in de meeste gevallen wordt aanbevolen.",
+ "Convert": "Omzetten",
+ "Export Audio": "Audio exporteren",
+ "Batch": "Batch",
+ "Input Folder": "Invoermap",
+ "Select the folder containing the audios to convert.": "Selecteer de map met de audio die u wilt converteren.",
+ "Enter input path": "Voer het invoerpad in",
+ "Output Folder": "Uitvoer map",
+ "Select the folder where the output audios will be saved.": "Selecteer de map waar de uitvoeraudio wordt opgeslagen.",
+ "Enter output path": "Voer het uitvoerpad in",
+ "Get information about the audio": "Informatie over de audio opvragen",
+ "Information about the audio file": "Informatie over het audiobestand",
+ "Waiting for information...": "Wachten op informatie...",
+ "## Voice Blender": "## Stem Blender",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Selecteer twee stemmodellen, stel het gewenste mengpercentage in en meng ze tot een geheel nieuwe stem.",
+ "Voice Blender": "Stem Blender",
+ "Drag and drop your model here": "Sleep uw model hierheen",
+ "You can also use a custom path.": "U kunt ook een aangepast pad gebruiken.",
+ "Blend Ratio": "Mengverhouding",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Door de positie meer naar de ene of de andere kant aan te passen, lijkt het model meer op het eerste of tweede.",
+ "Fusion": "Samensmelting",
+ "Path to Model": "Pad naar model",
+ "Enter path to model": "Pad naar model invoeren",
+ "Model information to be placed": "Te plaatsen modelinformatie",
+ "Inroduce the model information": "Produceer de modelinformatie",
+ "The information to be placed in the model (You can leave it blank or put anything).": "De informatie die in het model moet worden geplaatst (u kunt het leeg laten of alles plaatsen).",
+ "View model information": "Modelinformatie weergeven",
+ "Introduce the model pth path": "Introduceer het model pth pad",
+ "View": "Bekijken",
+ "Model extraction": "Extractie van modellen",
+ "Model conversion": "Model conversie",
+ "Pth file": "Pth-bestand",
+ "Output of the pth file": "Uitvoer van het pth-bestand",
+ "# How to Report an Issue on GitHub": "# Een probleem melden op GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Klik op de knop 'Opnamescherm' hieronder om te beginnen met het opnemen van het probleem dat u ondervindt.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Als u klaar bent met het opnemen van het probleem, klikt u op de knop 'Opname stoppen' (dezelfde knop, maar het label verandert afhankelijk van of u actief aan het opnemen bent of niet).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Ga naar [GitHub Issues](https://github.com/IAHispano/Applio/issues) en klik op de knop 'New Issue'.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Vul het meegeleverde uitgiftesjabloon in, zorg ervoor dat u indien nodig details opneemt en gebruik het gedeelte Activa om het opgenomen bestand uit de vorige stap te uploaden.",
+ "Record Screen": "Scherm opnemen",
+ "Record": "Record",
+ "Stop Recording": "Opname stoppen",
+ "Introduce the model .pth path": "Introduceer het model .pth-pad",
+ "See Model Information": "Modelinformatie bekijken",
+ "## Download Model": "## Model downloaden",
+ "Model Link": "Koppeling naar het model",
+ "Introduce the model link": "Introduceer de modellink",
+ "Download Model": "Model downloaden",
+ "## Drop files": "## Bestanden neerzetten",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Sleep uw .pth-bestand en .index-bestand naar deze ruimte. Sleep de ene en dan de andere.",
+ "TTS Voices": "TTS-stemmen",
+ "Select the TTS voice to use for the conversion.": "Selecteer de TTS-stem die u voor de conversie wilt gebruiken.",
+ "Text to Synthesize": "Tekst om te synthetiseren",
+ "Enter the text to synthesize.": "Voer de tekst in die u wilt synthetiseren.",
+ "Or you can upload a .txt file": "Of u kunt een .txt bestand uploaden",
+ "Enter text to synthesize": "Voer tekst in om te synthetiseren",
+ "Output Path for TTS Audio": "Uitvoerpad voor TTS-audio",
+ "Output Path for RVC Audio": "Uitvoerpad voor RVC-audio",
+ "Enable Applio integration with Discord presence": "Applio-integratie inschakelen met Discord-aanwezigheid",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Het activeert de mogelijkheid om de huidige Applio-activiteit in Discord weer te geven.",
+ "Enable Applio integration with applio.org/models using flask": "Applio-integratie met applio.org/models inschakelen met behulp van kolf",
+ "It will activate the possibility of downloading models with a click from the website.": "Het activeert de mogelijkheid om modellen te downloaden met een klik van de website.",
+ "Theme": "Thema",
+ "Select the theme you want to use. (Requires restarting Applio)": "Selecteer het thema dat je wilt gebruiken. (Vereist het opnieuw opstarten van Applio)",
+ "Language": "Taal",
+ "Select the language you want to use. (Requires restarting Applio)": "Selecteer de taal die u wilt gebruiken. (Vereist het opnieuw opstarten van Applio)",
+ "Plugin Installer": "Plug-in-installatieprogramma",
+ "Drag your plugin.zip to install it": "Sleep uw plugin.zip om deze te installeren",
+ "Version Checker": "Versie Checker",
+ "Check which version of Applio is the latest to see if you need to update.": "Controleer welke versie van Applio de nieuwste is om te zien of u moet updaten.",
+ "Check for updates": "Controleren op updates"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/pa_PA.json b/assets/i18n/languages/pa_PA.json
new file mode 100644
index 0000000000000000000000000000000000000000..249f3ea4e876288e83775ae5a43eb379c3a79899
--- /dev/null
+++ b/assets/i18n/languages/pa_PA.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "ਅੰਤਮ ਵੌਇਸ ਕਲੋਨਿੰਗ ਟੂਲ, ਬੇਮਿਸਾਲ ਸ਼ਕਤੀ, ਮਾਡਿਊਲਰਿਟੀ, ਅਤੇ ਉਪਭੋਗਤਾ-ਅਨੁਕੂਲ ਅਨੁਭਵ ਲਈ ਧਿਆਨ ਨਾਲ ਅਨੁਕੂਲ ਬਣਾਇਆ ਗਿਆ ਹੈ.",
+ "This section contains some extra utilities that often may be in experimental phases.": "ਇਸ ਭਾਗ ਵਿੱਚ ਕੁਝ ਵਾਧੂ ਉਪਯੋਗਤਾਵਾਂ ਹਨ ਜੋ ਅਕਸਰ ਪ੍ਰਯੋਗਾਤਮਕ ਪੜਾਵਾਂ ਵਿੱਚ ਹੋ ਸਕਦੀਆਂ ਹਨ।",
+ "Output Information": "ਆਊਟਪੁੱਟ ਜਾਣਕਾਰੀ",
+ "The output information will be displayed here.": "ਆਉਟਪੁੱਟ ਜਾਣਕਾਰੀ ਇੱਥੇ ਪ੍ਰਦਰਸ਼ਿਤ ਕੀਤੀ ਜਾਵੇਗੀ।",
+ "Inference": "ਅਨੁਮਾਨ",
+ "Train": "ਟ੍ਰੇਨ",
+ "Extra": "ਵਾਧੂ",
+ "Merge Audios": "ਆਡੀਓ ਨੂੰ ਮਿਲਾਓ",
+ "Processing": "ਪ੍ਰੋਸੈਸਿੰਗ",
+ "Audio Analyzer": "ਆਡੀਓ ਵਿਸ਼ਲੇਸ਼ਕ",
+ "Model Information": "ਮਾਡਲ ਜਾਣਕਾਰੀ",
+ "Plugins": "ਪਲੱਗਇਨ",
+ "Download": "ਡਾਊਨਲੋਡ ਕਰੋ",
+ "Report a Bug": "ਇੱਕ ਬਗ ਦੀ ਰਿਪੋਰਟ ਕਰੋ",
+ "Settings": "ਸੈਟਿੰਗਾਂ",
+ "Preprocess": "ਪ੍ਰੀਪ੍ਰੋਸੈਸ",
+ "Model Name": "ਮਾਡਲ ਦਾ ਨਾਮ",
+ "Name of the new model.": "ਨਵੇਂ ਮਾਡਲ ਦਾ ਨਾਮ।",
+ "Enter model name": "ਮਾਡਲ ਨਾਮ ਦਾਖਲ ਕਰੋ",
+ "Dataset Path": "ਡਾਟਾਸੈਟ ਪਾਥ",
+ "Path to the dataset folder.": "ਡੇਟਾਸੈਟ ਫੋਲਡਰ ਲਈ ਰਾਹ।",
+ "Refresh Datasets": "ਡੇਟਾਸੈਟਾਂ ਨੂੰ ਤਾਜ਼ਾ ਕਰੋ",
+ "Dataset Creator": "ਡਾਟਾਸੈਟ ਨਿਰਮਾਤਾ",
+ "Dataset Name": "ਡੇਟਾਸੈਟ ਨਾਮ",
+ "Name of the new dataset.": "ਨਵੇਂ ਡੇਟਾਸੈਟ ਦਾ ਨਾਮ।",
+ "Enter dataset name": "ਡੇਟਾਸੈਟ ਦਾ ਨਾਮ ਦਾਖਲ ਕਰੋ",
+ "Upload Audio Dataset": "ਆਡੀਓ ਡੇਟਾਸੈਟ ਅੱਪਲੋਡ ਕਰੋ",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ਆਡੀਓ ਫਾਇਲ ਨੂੰ ਡੇਟਾਸੈਟ ਵਿੱਚ ਸਫਲਤਾਪੂਰਵਕ ਜੋੜਿਆ ਗਿਆ ਹੈ। ਕਿਰਪਾ ਕਰਕੇ ਪ੍ਰੀਪ੍ਰੋਸੈਸ ਬਟਨ 'ਤੇ ਕਲਿੱਕ ਕਰੋ।",
+ "Enter dataset path": "ਡਾਟਾਸੈਟ ਪਾਥ ਦਾਖਲ ਕਰੋ",
+ "Sampling Rate": "ਨਮੂਨੇ ਲੈਣ ਦੀ ਦਰ",
+ "The sampling rate of the audio files.": "ਆਡੀਓ ਫਾਇਲਾਂ ਦੀ ਨਮੂਨੇ ਲੈਣ ਦੀ ਦਰ।",
+ "RVC Version": "RVC ਸੰਸਕਰਣ",
+ "The RVC version of the model.": "ਮਾਡਲ ਦਾ ਆਰਵੀਸੀ ਸੰਸਕਰਣ.",
+ "Preprocess Dataset": "ਪ੍ਰੀਪ੍ਰੋਸੈਸ ਡੇਟਾਸੈਟ",
+ "Extract": "ਐਕਸਟਰੈਕਟ",
+ "Hop Length": "ਹੌਪ ਲੰਬਾਈ",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "ਸਿਸਟਮ ਨੂੰ ਮਹੱਤਵਪੂਰਣ ਪਿਚ ਤਬਦੀਲੀ ਵਿੱਚ ਤਬਦੀਲ ਹੋਣ ਵਿੱਚ ਲੱਗਣ ਵਾਲੇ ਸਮੇਂ ਨੂੰ ਦਰਸਾਉਂਦਾ ਹੈ। ਛੋਟੀਆਂ ਹੌਪ ਲੰਬਾਈਆਂ ਨੂੰ ਅਨੁਮਾਨ ਲਗਾਉਣ ਲਈ ਵਧੇਰੇ ਸਮੇਂ ਦੀ ਲੋੜ ਹੁੰਦੀ ਹੈ ਪਰ ਪਿੱਚ ਦੀ ਸਟੀਕਤਾ ਵਧੇਰੇ ਹੁੰਦੀ ਹੈ।",
+ "Batch Size": "ਬੈਚ ਦਾ ਆਕਾਰ",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "ਇਸ ਨੂੰ ਆਪਣੇ GPU ਦੇ ਉਪਲਬਧ VRAM ਨਾਲ ਜੋੜਨ ਦੀ ਸਲਾਹ ਦਿੱਤੀ ਜਾਂਦੀ ਹੈ। 4 ਦੀ ਸੈਟਿੰਗ ਬਿਹਤਰ ਸ਼ੁੱਧਤਾ ਪਰ ਹੌਲੀ ਪ੍ਰੋਸੈਸਿੰਗ ਦੀ ਪੇਸ਼ਕਸ਼ ਕਰਦੀ ਹੈ, ਜਦੋਂ ਕਿ 8 ਤੇਜ਼ ਅਤੇ ਮਿਆਰੀ ਨਤੀਜੇ ਪ੍ਰਦਾਨ ਕਰਦੀ ਹੈ.",
+ "Save Every Epoch": "ਹਰ ਯੁੱਗ ਨੂੰ ਸੁਰੱਖਿਅਤ ਕਰੋ",
+ "Determine at how many epochs the model will saved at.": "ਇਹ ਨਿਰਧਾਰਤ ਕਰੋ ਕਿ ਮਾਡਲ ਕਿੰਨੇ ਯੁੱਗਾਂ ਵਿੱਚ ਸੁਰੱਖਿਅਤ ਹੋਵੇਗਾ।",
+ "Total Epoch": "ਕੁੱਲ ਯੁੱਗ",
+ "Specifies the overall quantity of epochs for the model training process.": "ਮਾਡਲ ਸਿਖਲਾਈ ਪ੍ਰਕਿਰਿਆ ਲਈ ਯੁੱਗਾਂ ਦੀ ਸਮੁੱਚੀ ਮਾਤਰਾ ਨਿਰਧਾਰਤ ਕਰਦਾ ਹੈ.",
+ "Pretrained": "ਪਹਿਲਾਂ ਤੋਂ ਸਿਖਲਾਈ ਪ੍ਰਾਪਤ",
+ "Save Only Latest": "ਕੇਵਲ ਨਵੀਨਤਮ ਨੂੰ ਸੁਰੱਖਿਅਤ ਕਰੋ",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "ਇਸ ਸੈਟਿੰਗ ਨੂੰ ਸਮਰੱਥ ਕਰਨ ਦੇ ਨਤੀਜੇ ਵਜੋਂ ਜੀ ਅਤੇ ਡੀ ਫਾਈਲਾਂ ਸਿਰਫ ਆਪਣੇ ਸਭ ਤੋਂ ਤਾਜ਼ਾ ਸੰਸਕਰਣਾਂ ਨੂੰ ਸੁਰੱਖਿਅਤ ਕਰਨਗੀਆਂ, ਸਟੋਰੇਜ ਸਪੇਸ ਨੂੰ ਪ੍ਰਭਾਵਸ਼ਾਲੀ ਢੰਗ ਨਾਲ ਸੁਰੱਖਿਅਤ ਕਰਨਗੀਆਂ.",
+ "Save Every Weights": "ਹਰ ਭਾਰ ਨੂੰ ਬਚਾਓ",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "ਇਹ ਸੈਟਿੰਗ ਤੁਹਾਨੂੰ ਹਰੇਕ ਯੁੱਗ ਦੇ ਅੰਤ 'ਤੇ ਮਾਡਲ ਦੇ ਭਾਰ ਨੂੰ ਬਚਾਉਣ ਦੇ ਯੋਗ ਬਣਾਉਂਦੀ ਹੈ।",
+ "Custom Pretrained": "ਕਸਟਮ ਪ੍ਰੀਟ੍ਰੇਨਡ",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "ਕਸਟਮ ਪ੍ਰੀਟ੍ਰੇਨਡ ਮਾਡਲਾਂ ਦੀ ਵਰਤੋਂ ਕਰਨ ਨਾਲ ਵਧੀਆ ਨਤੀਜੇ ਨਿਕਲ ਸਕਦੇ ਹਨ, ਕਿਉਂਕਿ ਵਿਸ਼ੇਸ਼ ਵਰਤੋਂ ਦੇ ਕੇਸ ਦੇ ਅਨੁਸਾਰ ਸਭ ਤੋਂ ਢੁਕਵੇਂ ਪੂਰਵ-ਸਿਖਲਾਈ ਪ੍ਰਾਪਤ ਮਾਡਲਾਂ ਦੀ ਚੋਣ ਕਰਨ ਨਾਲ ਪ੍ਰਦਰਸ਼ਨ ਵਿੱਚ ਮਹੱਤਵਪੂਰਣ ਵਾਧਾ ਹੋ ਸਕਦਾ ਹੈ.",
+ "Upload Pretrained Model": "ਪਹਿਲਾਂ ਤੋਂ ਸਿਖਲਾਈ ਪ੍ਰਾਪਤ ਮਾਡਲ ਅਪਲੋਡ ਕਰੋ",
+ "Refresh Custom Pretraineds": "ਕਸਟਮ ਪ੍ਰੀਟ੍ਰੇਨਡਾਂ ਨੂੰ ਤਾਜ਼ਾ ਕਰੋ",
+ "Pretrained Custom Settings": "ਪਹਿਲਾਂ ਤੋਂ ਸਿਖਲਾਈ ਪ੍ਰਾਪਤ ਕਸਟਮ ਸੈਟਿੰਗਾਂ",
+ "The file you dropped is not a valid pretrained file. Please try again.": "ਤੁਹਾਡੇ ਵੱਲੋਂ ਛੱਡੀ ਗਈ ਫਾਇਲ ਇੱਕ ਵੈਧ ਪੂਰਵ-ਸਿਖਲਾਈ ਪ੍ਰਾਪਤ ਫਾਇਲ ਨਹੀਂ ਹੈ। ਕਿਰਪਾ ਕਰਕੇ ਦੁਬਾਰਾ ਕੋਸ਼ਿਸ਼ ਕਰੋ।",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "ਡ੍ਰੌਪਡਾਊਨ ਮੀਨੂ ਵਿੱਚ ਪਹਿਲਾਂ ਤੋਂ ਸਿਖਲਾਈ ਪ੍ਰਾਪਤ ਫਾਇਲ ਨੂੰ ਦੇਖਣ ਲਈ ਰੀਫਰੈਸ਼ ਬਟਨ 'ਤੇ ਕਲਿੱਕ ਕਰੋ।",
+ "Pretrained G Path": "ਕਸਟਮ ਪ੍ਰੀਟ੍ਰੇਨਡ G",
+ "Pretrained D Path": "ਕਸਟਮ ਪ੍ਰੀਟ੍ਰੇਨਡ ਡੀ",
+ "GPU Settings": "GPU ਸੈਟਿੰਗਾਂ",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "ਉੱਨਤ GPU ਸੈਟਿੰਗਾਂ ਸੈੱਟ ਕਰਦਾ ਹੈ, ਜੋ ਬਿਹਤਰ GPU ਆਰਕੀਟੈਕਚਰ ਵਾਲੇ ਉਪਭੋਗਤਾਵਾਂ ਲਈ ਸਿਫਾਰਸ਼ ਕੀਤੀ ਜਾਂਦੀ ਹੈ।",
+ "GPU Custom Settings": "GPU ਕਸਟਮ ਸੈਟਿੰਗਾਂ",
+ "GPU Number": "GPU ਨੰਬਰ",
+ "0 to ∞ separated by -": "0 ਤੋਂ ∞ ਦੁਆਰਾ ਵੱਖ ਕੀਤਾ ਜਾਂਦਾ ਹੈ -",
+ "GPU Information": "GPU ਜਾਣਕਾਰੀ",
+ "Pitch Guidance": "ਪਿਚ ਗਾਈਡੈਂਸ",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "ਪਿਚ ਗਾਈਡੈਂਸ ਦੀ ਵਰਤੋਂ ਕਰਕੇ, ਅਸਲ ਆਵਾਜ਼ ਦੇ ਇੰਟਰਨੇਸ਼ਨ ਨੂੰ ਦਰਸਾਉਣਾ ਸੰਭਵ ਹੋ ਜਾਂਦਾ ਹੈ, ਜਿਸ ਵਿੱਚ ਇਸਦੀ ਪਿੱਚ ਵੀ ਸ਼ਾਮਲ ਹੈ. ਇਹ ਵਿਸ਼ੇਸ਼ਤਾ ਗਾਇਕੀ ਅਤੇ ਹੋਰ ਦ੍ਰਿਸ਼ਾਂ ਲਈ ਵਿਸ਼ੇਸ਼ ਤੌਰ 'ਤੇ ਕੀਮਤੀ ਹੈ ਜਿੱਥੇ ਮੂਲ ਸੁਰ ਜਾਂ ਪਿੱਚ ਪੈਟਰਨ ਨੂੰ ਸੁਰੱਖਿਅਤ ਰੱਖਣਾ ਜ਼ਰੂਰੀ ਹੈ।",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "ਆਪਣੇ ਆਪ ਨੂੰ ਸਿਖਲਾਈ ਦਿੰਦੇ ਸਮੇਂ ਪਹਿਲਾਂ ਤੋਂ ਸਿਖਲਾਈ ਪ੍ਰਾਪਤ ਮਾਡਲਾਂ ਦੀ ਵਰਤੋਂ ਕਰੋ। ਇਹ ਪਹੁੰਚ ਸਿਖਲਾਈ ਦੀ ਮਿਆਦ ਨੂੰ ਘਟਾਉਂਦੀ ਹੈ ਅਤੇ ਸਮੁੱਚੀ ਗੁਣਵੱਤਾ ਨੂੰ ਵਧਾਉਂਦੀ ਹੈ।",
+ "Extract Features": "ਐਕਸਟਰੈਕਟ ਵਿਸ਼ੇਸ਼ਤਾਵਾਂ",
+ "Start Training": "ਸਿਖਲਾਈ ਸ਼ੁਰੂ ਕਰੋ",
+ "Generate Index": "ਇੰਡੈਕਸ ਜਨਰੇਟ ਕਰੋ",
+ "Voice Model": "ਵੌਇਸ ਮਾਡਲ",
+ "Select the voice model to use for the conversion.": "ਪਰਿਵਰਤਨ ਲਈ ਵਰਤਣ ਲਈ ਵੌਇਸ ਮਾਡਲ ਦੀ ਚੋਣ ਕਰੋ।",
+ "Index File": "ਇੰਡੈਕਸ ਫਾਇਲ",
+ "Select the index file to use for the conversion.": "ਪਰਿਵਰਤਨ ਲਈ ਵਰਤਣ ਲਈ ਇੰਡੈਕਸ ਫਾਇਲ ਦੀ ਚੋਣ ਕਰੋ।",
+ "Refresh": "ਤਾਜ਼ਾ ਕਰੋ",
+ "Unload Voice": "ਆਵਾਜ਼ ਨੂੰ ਅਨਲੋਡ ਕਰੋ",
+ "Single": "ਸਿੰਗਲ",
+ "Upload Audio": "ਆਡੀਓ ਅੱਪਲੋਡ ਕਰੋ",
+ "Select Audio": "ਆਡੀਓ ਚੁਣੋ",
+ "Select the audio to convert.": "ਕਨਵਰਟ ਕਰਨ ਲਈ ਆਡੀਓ ਦੀ ਚੋਣ ਕਰੋ।",
+ "Advanced Settings": "ਉੱਨਤ ਸੈਟਿੰਗਾਂ",
+ "Clear Outputs (Deletes all audios in assets/audios)": "ਸਪਸ਼ਟ ਆਊਟਪੁੱਟ (ਜਾਇਦਾਦਾਂ/ਆਡੀਓ ਵਿੱਚ ਸਾਰੇ ਆਡੀਓ ਮਿਟਾ ਦਿੰਦਾ ਹੈ)",
+ "Custom Output Path": "ਕਸਟਮ ਆਉਟਪੁੱਟ ਪਾਥ",
+ "Output Path": "ਆਊਟਪੁੱਟ ਪਾਥ",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "ਉਹ ਰਸਤਾ ਜਿੱਥੇ ਆਉਟਪੁੱਟ ਆਡੀਓ ਨੂੰ ਸੁਰੱਖਿਅਤ ਕੀਤਾ ਜਾਵੇਗਾ, ਸੰਪਤੀਆਂ/ਆਡੀਓਜ਼/output.wav",
+ "Split Audio": "ਸਪਲਿਟ ਆਡੀਓ",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "ਕੁਝ ਮਾਮਲਿਆਂ ਵਿੱਚ ਬਿਹਤਰ ਨਤੀਜੇ ਪ੍ਰਾਪਤ ਕਰਨ ਲਈ ਅਨੁਮਾਨ ਲਈ ਆਡੀਓ ਨੂੰ ਟੁਕੜਿਆਂ ਵਿੱਚ ਵੰਡੋ।",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "ਆਪਣੇ ਅਨੁਮਾਨਾਂ ਲਈ ਇੱਕ ਨਰਮ ਆਟੋਟਿਊਨ ਲਾਗੂ ਕਰੋ, ਜਿਸ ਦੀ ਸਿਫਾਰਸ਼ ਗਾਇਨ ਪਰਿਵਰਤਨ ਲਈ ਕੀਤੀ ਜਾਂਦੀ ਹੈ।",
+ "Clean Audio": "ਕਲੀਨ ਆਡੀਓ",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "ਆਡੀਓ ਬੋਲਣ ਲਈ ਸਿਫਾਰਸ਼ ਕੀਤੇ ਗਏ ਸ਼ੋਰ ਦਾ ਪਤਾ ਲਗਾਉਣ ਵਾਲੇ ਐਲਗੋਰਿਦਮ ਦੀ ਵਰਤੋਂ ਕਰਕੇ ਆਪਣੇ ਆਡੀਓ ਆਉਟਪੁੱਟ ਨੂੰ ਸਾਫ਼ ਕਰੋ।",
+ "Clean Strength": "ਸਾਫ਼ ਤਾਕਤ",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "ਜਿਸ ਆਡੀਓ ਨੂੰ ਤੁਸੀਂ ਚਾਹੁੰਦੇ ਹੋ ਉਸ 'ਤੇ ਕਲੀਨ-ਅੱਪ ਪੱਧਰ ਸੈੱਟ ਕਰੋ, ਜਿੰਨਾ ਜ਼ਿਆਦਾ ਤੁਸੀਂ ਇਸ ਨੂੰ ਵਧਾਓਗੇ, ਓਨਾ ਹੀ ਇਹ ਸਾਫ਼ ਹੋ ਜਾਵੇਗਾ, ਪਰ ਇਹ ਸੰਭਵ ਹੈ ਕਿ ਆਡੀਓ ਵਧੇਰੇ ਸੰਕੁਚਿਤ ਹੋ ਜਾਵੇਗਾ.",
+ "Pitch": "ਪਿਚ",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ਆਡੀਓ ਦੀ ਪਿਚ ਸੈੱਟ ਕਰੋ, ਜਿੰਨਾ ਜ਼ਿਆਦਾ ਮੁੱਲ ਹੋਵੇਗਾ, ਪਿਚ ਓਨੀ ਹੀ ਉੱਚੀ ਹੋਵੇਗੀ।",
+ "Filter Radius": "ਫਿਲਟਰ ਰੇਡੀਅਸ",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "ਜੇ ਗਿਣਤੀ ਤਿੰਨ ਤੋਂ ਵੱਧ ਜਾਂ ਬਰਾਬਰ ਹੈ, ਤਾਂ ਇਕੱਤਰ ਕੀਤੇ ਟੋਨ ਨਤੀਜਿਆਂ 'ਤੇ ਮੀਡੀਅਨ ਫਿਲਟਰਿੰਗ ਲਗਾਉਣ ਨਾਲ ਸਾਹ ਲੈਣ ਵਿੱਚ ਕਮੀ ਆਉਣ ਦੀ ਸੰਭਾਵਨਾ ਹੁੰਦੀ ਹੈ.",
+ "Search Feature Ratio": "ਖੋਜ ਵਿਸ਼ੇਸ਼ਤਾ ਅਨੁਪਾਤ",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "ਇੰਡੈਕਸ ਫਾਈਲ ਦੁਆਰਾ ਪਾਇਆ ਗਿਆ ਪ੍ਰਭਾਵ; ਇੱਕ ਉੱਚ ਮੁੱਲ ਵਧੇਰੇ ਪ੍ਰਭਾਵ ਨਾਲ ਮੇਲ ਖਾਂਦਾ ਹੈ. ਹਾਲਾਂਕਿ, ਘੱਟ ਮੁੱਲਾਂ ਦੀ ਚੋਣ ਕਰਨਾ ਆਡੀਓ ਵਿੱਚ ਮੌਜੂਦ ਕਲਾਕ੍ਰਿਤੀਆਂ ਨੂੰ ਘਟਾਉਣ ਵਿੱਚ ਮਦਦ ਕਰ ਸਕਦਾ ਹੈ।",
+ "Volume Envelope": "ਵਾਲਿਊਮ ਲਿਫਾਫਾ",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "ਆਉਟਪੁੱਟ ਦੇ ਵਾਲੀਅਮ ਲਿਫਾਫੇ ਨੂੰ ਬਦਲੋ ਜਾਂ ਮਿਸ਼ਰਣ ਕਰੋ। ਅਨੁਪਾਤ ਜਿੰਨਾ 1 ਦੇ ਨੇੜੇ ਹੁੰਦਾ ਹੈ, ਓਨਾ ਹੀ ਆਉਟਪੁੱਟ ਲਿਫਾਫਾ ਵਰਤਿਆ ਜਾਂਦਾ ਹੈ.",
+ "Protect Voiceless Consonants": "ਅਵਾਜ਼ ਰਹਿਤ ਵਿੰਜਨਾਂ ਦੀ ਰੱਖਿਆ ਕਰੋ",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "ਇਲੈਕਟ੍ਰੋ-ਐਕੋਸਟਿਕ ਫਾੜਨ ਅਤੇ ਹੋਰ ਕਲਾਕ੍ਰਿਤੀਆਂ ਨੂੰ ਰੋਕਣ ਲਈ ਵੱਖਰੇ ਵਿਜਨਾਂ ਅਤੇ ਸਾਹ ਲੈਣ ਦੀਆਂ ਆਵਾਜ਼ਾਂ ਦੀ ਰੱਖਿਆ ਕਰੋ। ਪੈਰਾਮੀਟਰ ਨੂੰ ਇਸਦੇ ਵੱਧ ਤੋਂ ਵੱਧ ਮੁੱਲ 0.5 ਤੱਕ ਖਿੱਚਣਾ ਵਿਆਪਕ ਸੁਰੱਖਿਆ ਪ੍ਰਦਾਨ ਕਰਦਾ ਹੈ। ਹਾਲਾਂਕਿ, ਇਸ ਮੁੱਲ ਨੂੰ ਘਟਾਉਣ ਨਾਲ ਸੁਰੱਖਿਆ ਦੀ ਹੱਦ ਘੱਟ ਹੋ ਸਕਦੀ ਹੈ ਜਦੋਂ ਕਿ ਸੰਭਾਵਿਤ ਤੌਰ 'ਤੇ ਸੂਚਕਪ੍ਰਭਾਵ ਨੂੰ ਘਟਾਇਆ ਜਾ ਸਕਦਾ ਹੈ.",
+ "Pitch extraction algorithm": "ਪਿਚ ਐਕਸਟਰੈਕਸ਼ਨ ਐਲਗੋਰਿਦਮ",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ਆਡੀਓ ਪਰਿਵਰਤਨ ਲਈ ਵਰਤਣ ਲਈ ਪਿਚ ਐਕਸਟਰੈਕਸ਼ਨ ਐਲਗੋਰਿਦਮ. ਡਿਫਾਲਟ ਐਲਗੋਰਿਦਮ rmvpe ਹੈ, ਜਿਸਦੀ ਸਿਫਾਰਸ਼ ਜ਼ਿਆਦਾਤਰ ਮਾਮਲਿਆਂ ਲਈ ਕੀਤੀ ਜਾਂਦੀ ਹੈ।",
+ "Convert": "ਕਨਵਰਟ ਕਰੋ",
+ "Export Audio": "ਆਡੀਓ ਨਿਰਯਾਤ ਕਰੋ",
+ "Batch": "ਬੈਚ",
+ "Input Folder": "ਇਨਪੁੱਟ ਫੋਲਡਰ",
+ "Select the folder containing the audios to convert.": "ਬਦਲਣ ਲਈ ਆਡੀਓ ਵਾਲੇ ਫੋਲਡਰ ਦੀ ਚੋਣ ਕਰੋ।",
+ "Enter input path": "ਇਨਪੁੱਟ ਪਾਥ ਦਾਖਲ ਕਰੋ",
+ "Output Folder": "ਆਊਟਪੁੱਟ ਫੋਲਡਰ",
+ "Select the folder where the output audios will be saved.": "ਉਹ ਫੋਲਡਰ ਚੁਣੋ ਜਿੱਥੇ ਆਉਟਪੁੱਟ ਆਡੀਓ ਸੁਰੱਖਿਅਤ ਕੀਤੇ ਜਾਣਗੇ।",
+ "Enter output path": "ਆਊਟਪੁੱਟ ਪਾਥ ਦਾਖਲ ਕਰੋ",
+ "Get information about the audio": "ਆਡੀਓ ਬਾਰੇ ਜਾਣਕਾਰੀ ਪ੍ਰਾਪਤ ਕਰੋ",
+ "Information about the audio file": "ਆਡੀਓ ਫਾਇਲ ਬਾਰੇ ਜਾਣਕਾਰੀ",
+ "Waiting for information...": "ਜਾਣਕਾਰੀ ਦੀ ਉਡੀਕ ਕੀਤੀ ਜਾ ਰਹੀ ਹੈ...",
+ "## Voice Blender": "## ਵੌਇਸ ਬਲੇਂਡਰ",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "ਦੋ ਵੌਇਸ ਮਾਡਲਾਂ ਦੀ ਚੋਣ ਕਰੋ, ਆਪਣੀ ਲੋੜੀਂਦੀ ਮਿਸ਼ਰਣ ਪ੍ਰਤੀਸ਼ਤਤਾ ਸੈੱਟ ਕਰੋ, ਅਤੇ ਉਨ੍ਹਾਂ ਨੂੰ ਪੂਰੀ ਤਰ੍ਹਾਂ ਨਵੀਂ ਆਵਾਜ਼ ਵਿੱਚ ਮਿਲਾਓ.",
+ "Voice Blender": "ਵੌਇਸ ਬਲੇਂਡਰ",
+ "Drag and drop your model here": "ਆਪਣੇ ਮਾਡਲ ਨੂੰ ਇੱਥੇ ਖਿੱਚ ਕੇ ਛੱਡ ਦਿਓ",
+ "You can also use a custom path.": "ਤੁਸੀਂ ਇੱਕ ਕਸਟਮ ਪਾਥ ਦੀ ਵਰਤੋਂ ਵੀ ਕਰ ਸਕਦੇ ਹੋ।",
+ "Blend Ratio": "ਮਿਸ਼ਰਣ ਅਨੁਪਾਤ",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "ਸਥਿਤੀ ਨੂੰ ਇੱਕ ਪਾਸੇ ਜਾਂ ਦੂਜੇ ਪਾਸੇ ਵਧੇਰੇ ਵਿਵਸਥਿਤ ਕਰਨਾ ਮਾਡਲ ਨੂੰ ਪਹਿਲੇ ਜਾਂ ਦੂਜੇ ਦੇ ਸਮਾਨ ਬਣਾ ਦੇਵੇਗਾ।",
+ "Fusion": "ਫਿਊਜ਼ਨ",
+ "Path to Model": "ਮਾਡਲ ਲਈ ਰਾਹ",
+ "Enter path to model": "ਮਾਡਲ ਲਈ ਪਾਥ ਦਾਖਲ ਕਰੋ",
+ "Model information to be placed": "ਮਾਡਲ ਜਾਣਕਾਰੀ ਰੱਖੀ ਜਾਣੀ ਚਾਹੀਦੀ ਹੈ",
+ "Inroduce the model information": "ਮਾਡਲ ਜਾਣਕਾਰੀ ਨੂੰ ਪੇਸ਼ ਕਰੋ",
+ "The information to be placed in the model (You can leave it blank or put anything).": "ਮਾਡਲ ਵਿੱਚ ਰੱਖੀ ਜਾਣ ਵਾਲੀ ਜਾਣਕਾਰੀ (ਤੁਸੀਂ ਇਸ ਨੂੰ ਖਾਲੀ ਛੱਡ ਸਕਦੇ ਹੋ ਜਾਂ ਕੁਝ ਵੀ ਪਾ ਸਕਦੇ ਹੋ)।",
+ "View model information": "ਮਾਡਲ ਜਾਣਕਾਰੀ ਦੇਖੋ",
+ "Introduce the model pth path": "ਮਾਡਲ pth ਪਾਥ ਨੂੰ ਪੇਸ਼ ਕਰੋ",
+ "View": "ਦ੍ਰਿਸ਼",
+ "Model extraction": "ਮਾਡਲ ਕੱਢਣਾ",
+ "Model conversion": "ਮਾਡਲ ਪਰਿਵਰਤਨ",
+ "Pth file": "Pth ਫਾਇਲ",
+ "Output of the pth file": "pth ਫਾਇਲ ਦਾ ਆਊਟਪੁੱਟ",
+ "# How to Report an Issue on GitHub": "# GitHub 'ਤੇ ਕਿਸੇ ਮੁੱਦੇ ਦੀ ਰਿਪੋਰਟ ਕਿਵੇਂ ਕਰਨੀ ਹੈ",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. ਤੁਹਾਡੇ ਵੱਲੋਂ ਅਨੁਭਵ ਕੀਤੀ ਜਾ ਰਹੀ ਸਮੱਸਿਆ ਨੂੰ ਰਿਕਾਰਡ ਕਰਨਾ ਸ਼ੁਰੂ ਕਰਨ ਲਈ ਹੇਠਾਂ ਦਿੱਤੇ 'ਰਿਕਾਰਡ ਸਕ੍ਰੀਨ' ਬਟਨ 'ਤੇ ਕਲਿੱਕ ਕਰੋ।",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. ਇੱਕ ਵਾਰ ਜਦੋਂ ਤੁਸੀਂ ਸਮੱਸਿਆ ਨੂੰ ਰਿਕਾਰਡ ਕਰਨਾ ਪੂਰਾ ਕਰ ਲੈਂਦੇ ਹੋ, ਤਾਂ 'ਸਟਾਪ ਰਿਕਾਰਡਿੰਗ' ਬਟਨ 'ਤੇ ਕਲਿੱਕ ਕਰੋ (ਉਹੀ ਬਟਨ, ਪਰ ਲੇਬਲ ਇਸ ਗੱਲ 'ਤੇ ਨਿਰਭਰ ਕਰਦਾ ਹੈ ਕਿ ਤੁਸੀਂ ਸਰਗਰਮੀ ਨਾਲ ਰਿਕਾਰਡਿੰਗ ਕਰ ਰਹੇ ਹੋ ਜਾਂ ਨਹੀਂ)।",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub ਮੁੱਦੇ] (https://github.com/IAHispano/Applio/issues) 'ਤੇ ਜਾਓ ਅਤੇ 'ਨਵਾਂ ਮੁੱਦਾ' ਬਟਨ 'ਤੇ ਕਲਿੱਕ ਕਰੋ।",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. ਪ੍ਰਦਾਨ ਕੀਤੇ ਮੁੱਦੇ ਦੇ ਟੈਂਪਲੇਟ ਨੂੰ ਪੂਰਾ ਕਰੋ, ਲੋੜ ਅਨੁਸਾਰ ਵੇਰਵੇ ਸ਼ਾਮਲ ਕਰਨਾ ਯਕੀਨੀ ਬਣਾਓ, ਅਤੇ ਪਿਛਲੇ ਕਦਮ ਤੋਂ ਰਿਕਾਰਡ ਕੀਤੀ ਫਾਈਲ ਨੂੰ ਅੱਪਲੋਡ ਕਰਨ ਲਈ ਸੰਪਤੀ ਸੈਕਸ਼ਨ ਦੀ ਵਰਤੋਂ ਕਰੋ.",
+ "Record Screen": "ਰਿਕਾਰਡ ਸਕ੍ਰੀਨ",
+ "Record": "ਰਿਕਾਰਡ",
+ "Stop Recording": "ਰਿਕਾਰਡਿੰਗ ਬੰਦ ਕਰੋ",
+ "Introduce the model .pth path": "ਮਾਡਲ .pth path ਨੂੰ ਪੇਸ਼ ਕਰੋ",
+ "See Model Information": "ਮਾਡਲ ਜਾਣਕਾਰੀ ਦੇਖੋ",
+ "## Download Model": "## ਡਾਊਨਲੋਡ ਮਾਡਲ",
+ "Model Link": "ਮਾਡਲ ਲਿੰਕ",
+ "Introduce the model link": "ਮਾਡਲ ਲਿੰਕ ਪੇਸ਼ ਕਰੋ",
+ "Download Model": "ਮਾਡਲ ਡਾਊਨਲੋਡ ਕਰੋ",
+ "## Drop files": "## ਫਾਇਲਾਂ ਛੱਡੋ",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "ਆਪਣੀ .pth ਫਾਇਲ ਅਤੇ .index ਫਾਇਲ ਨੂੰ ਇਸ ਸਪੇਸ ਵਿੱਚ ਖਿੱਚੋ। ਇੱਕ ਨੂੰ ਖਿੱਚੋ ਅਤੇ ਫਿਰ ਦੂਜੇ ਨੂੰ।",
+ "TTS Voices": "TTS ਆਵਾਜ਼ਾਂ",
+ "Select the TTS voice to use for the conversion.": "ਪਰਿਵਰਤਨ ਲਈ ਵਰਤਣ ਲਈ TTS ਆਵਾਜ਼ ਦੀ ਚੋਣ ਕਰੋ।",
+ "Text to Synthesize": "ਸੰਸ਼ਲੇਸ਼ਣ ਕਰਨ ਲਈ ਟੈਕਸਟ",
+ "Enter the text to synthesize.": "ਸੰਸ਼ਲੇਸ਼ਿਤ ਕਰਨ ਲਈ ਟੈਕਸਟ ਦਾਖਲ ਕਰੋ।",
+ "Or you can upload a .txt file": "ਜਾਂ ਤੁਸੀਂ ਕੋਈ .txt ਫਾਇਲ ਅੱਪਲੋਡ ਕਰ ਸਕਦੇ ਹੋ",
+ "Enter text to synthesize": "ਸੰਸ਼ਲੇਸ਼ਿਤ ਕਰਨ ਲਈ ਟੈਕਸਟ ਦਾਖਲ ਕਰੋ",
+ "Output Path for TTS Audio": "TTS ਆਡੀਓ ਲਈ ਆਉਟਪੁੱਟ ਪਾਥ",
+ "Output Path for RVC Audio": "RVC ਆਡੀਓ ਲਈ ਆਉਟਪੁੱਟ ਪਾਥ",
+ "Enable Applio integration with Discord presence": "ਡਿਸਕਾਰਡ ਦੀ ਮੌਜੂਦਗੀ ਨਾਲ ਐਪਲੀਓ ਏਕੀਕਰਣ ਨੂੰ ਸਮਰੱਥ ਕਰੋ",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "ਇਹ ਡਿਸਕਾਰਡ ਵਿੱਚ ਮੌਜੂਦਾ ਐਪਲੀਓ ਗਤੀਵਿਧੀ ਨੂੰ ਪ੍ਰਦਰਸ਼ਿਤ ਕਰਨ ਦੀ ਸੰਭਾਵਨਾ ਨੂੰ ਕਿਰਿਆਸ਼ੀਲ ਕਰੇਗਾ।",
+ "Enable Applio integration with applio.org/models using flask": "ਫਲਾਸਕ ਦੀ ਵਰਤੋਂ ਕਰਕੇ applio.org/models ਨਾਲ ਐਪਲਿਓ ਏਕੀਕਰਣ ਨੂੰ ਸਮਰੱਥ ਕਰੋ",
+ "It will activate the possibility of downloading models with a click from the website.": "ਇਹ ਵੈਬਸਾਈਟ ਤੋਂ ਇੱਕ ਕਲਿੱਕ ਨਾਲ ਮਾਡਲਾਂ ਨੂੰ ਡਾਊਨਲੋਡ ਕਰਨ ਦੀ ਸੰਭਾਵਨਾ ਨੂੰ ਕਿਰਿਆਸ਼ੀਲ ਕਰੇਗਾ।",
+ "Theme": "ਥੀਮ",
+ "Select the theme you want to use. (Requires restarting Applio)": "ਉਹ ਥੀਮ ਚੁਣੋ ਜਿਸਨੂੰ ਤੁਸੀਂ ਵਰਤਣਾ ਚਾਹੁੰਦੇ ਹੋ। (ਐਪਲੀਕੇਸ਼ਨ ਨੂੰ ਦੁਬਾਰਾ ਸ਼ੁਰੂ ਕਰਨ ਦੀ ਲੋੜ ਹੈ)",
+ "Language": "ਭਾਸ਼ਾ",
+ "Select the language you want to use. (Requires restarting Applio)": "ਉਹ ਭਾਸ਼ਾ ਚੁਣੋ ਜਿਸਨੂੰ ਤੁਸੀਂ ਵਰਤਣਾ ਚਾਹੁੰਦੇ ਹੋ। (ਐਪਲੀਕੇਸ਼ਨ ਨੂੰ ਦੁਬਾਰਾ ਸ਼ੁਰੂ ਕਰਨ ਦੀ ਲੋੜ ਹੈ)",
+ "Plugin Installer": "ਪਲੱਗਇਨ ਇੰਸਟਾਲਰ",
+ "Drag your plugin.zip to install it": "ਇਸ ਨੂੰ ਇੰਸਟਾਲ ਕਰਨ ਲਈ ਆਪਣੇ plugin.zip ਨੂੰ ਖਿੱਚੋ",
+ "Version Checker": "ਸੰਸਕਰਣ ਚੈਕਰ",
+ "Check which version of Applio is the latest to see if you need to update.": "ਇਹ ਦੇਖਣ ਲਈ ਜਾਂਚ ਕਰੋ ਕਿ ਕੀ ਤੁਹਾਨੂੰ ਅੱਪਡੇਟ ਕਰਨ ਦੀ ਲੋੜ ਹੈ, ਐਪਲੀਓ ਦਾ ਕਿਹੜਾ ਸੰਸਕਰਣ ਨਵੀਨਤਮ ਹੈ।",
+ "Check for updates": "ਅੱਪਡੇਟਾਂ ਵਾਸਤੇ ਜਾਂਚ ਕਰੋ"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/pl_PL.json b/assets/i18n/languages/pl_PL.json
new file mode 100644
index 0000000000000000000000000000000000000000..60399c1e1cf41b7b319a45f069f45a5537237be3
--- /dev/null
+++ b/assets/i18n/languages/pl_PL.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Najlepsze narzędzie do klonowania głosu, skrupulatnie zoptymalizowane pod kątem niezrównanej mocy, modułowości i przyjazności dla użytkownika.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Ta sekcja zawiera kilka dodatkowych narzędzi, które często mogą znajdować się w fazie eksperymentalnej.",
+ "Output Information": "Informacje wyjściowe",
+ "The output information will be displayed here.": "W tym miejscu zostaną wyświetlone informacje wyjściowe.",
+ "Inference": "Wnioskowanie",
+ "Train": "Pociąg",
+ "Extra": "Dodatkowych",
+ "Merge Audios": "Scal audio",
+ "Processing": "Przetwarzanie",
+ "Audio Analyzer": "Analizator dźwięku",
+ "Model Information": "Informacje o modelu",
+ "Plugins": "Wtyczki",
+ "Download": "Pobierać",
+ "Report a Bug": "Zgłoś błąd",
+ "Settings": "Ustawienia",
+ "Preprocess": "Przetwarzanie wstępne",
+ "Model Name": "Nazwa modelu",
+ "Name of the new model.": "Nazwa nowego modelu.",
+ "Enter model name": "Wprowadź nazwę modelu",
+ "Dataset Path": "Ścieżka zestawu danych",
+ "Path to the dataset folder.": "Ścieżka do folderu zestawu danych.",
+ "Refresh Datasets": "Odświeżanie zestawów danych",
+ "Dataset Creator": "Twórca zestawu danych",
+ "Dataset Name": "Nazwa zestawu danych",
+ "Name of the new dataset.": "Nazwa nowego zestawu danych.",
+ "Enter dataset name": "Wprowadź nazwę zestawu danych",
+ "Upload Audio Dataset": "Przekazywanie zestawu danych audio",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Plik audio został pomyślnie dodany do zestawu danych. Kliknij przycisk przetwarzania wstępnego.",
+ "Enter dataset path": "Wprowadź ścieżkę zestawu danych",
+ "Sampling Rate": "Częstotliwość próbkowania",
+ "The sampling rate of the audio files.": "Częstotliwość próbkowania plików audio.",
+ "RVC Version": "Wersja RVC",
+ "The RVC version of the model.": "Wersja modelu RVC.",
+ "Preprocess Dataset": "Wstępne przetwarzanie zestawu danych",
+ "Extract": "Ekstrakt",
+ "Hop Length": "Długość chmielu",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Oznacza czas potrzebny systemowi do przejścia do znaczącej zmiany wysokości dźwięku. Mniejsze długości przeskoków wymagają więcej czasu na wnioskowanie, ale zwykle zapewniają wyższą dokładność skoku.",
+ "Batch Size": "Wielkość partii",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Wskazane jest, aby dostosować go do dostępnej pamięci VRAM procesora graficznego. Ustawienie 4 zapewnia lepszą dokładność, ale wolniejsze przetwarzanie, podczas gdy 8 zapewnia szybsze i standardowe wyniki.",
+ "Save Every Epoch": "Ocal każdą epokę",
+ "Determine at how many epochs the model will saved at.": "Określ, w ilu epokach model zostanie zapisany.",
+ "Total Epoch": "Epoka ogółem",
+ "Specifies the overall quantity of epochs for the model training process.": "Określa całkowitą liczbę epok dla procesu trenowania modelu.",
+ "Pretrained": "Wstępnie wytrenowany",
+ "Save Only Latest": "Zapisz tylko najnowsze",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Włączenie tego ustawienia spowoduje, że pliki G i D będą zapisywać tylko swoje najnowsze wersje, skutecznie oszczędzając miejsce na dysku.",
+ "Save Every Weights": "Oszczędzaj wszystkie ciężary",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "To ustawienie umożliwia zapisanie wag modelu na końcu każdej epoki.",
+ "Custom Pretrained": "Niestandardowe wstępnie wytrenowane",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Korzystanie z niestandardowych wstępnie wytrenowanych modeli może prowadzić do doskonałych wyników, ponieważ wybranie najbardziej odpowiednich wstępnie wytrenowanych modeli dostosowanych do konkretnego przypadku użycia może znacznie zwiększyć wydajność.",
+ "Upload Pretrained Model": "Przekazywanie wstępnie wytrenowanego modelu",
+ "Refresh Custom Pretraineds": "Odświeżanie niestandardowych wstępnie wytrenowanych",
+ "Pretrained Custom Settings": "Wstępnie wytrenowane ustawienia niestandardowe",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Upuszczony plik nie jest prawidłowym wstępnie wytrenowanym plikiem. Spróbuj ponownie.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Kliknij przycisk odświeżania, aby wyświetlić wstępnie wytrenowany plik w menu rozwijanym.",
+ "Pretrained G Path": "Niestandardowe wstępnie wytrenowane G",
+ "Pretrained D Path": "Niestandardowy wstępnie wytrenowany D",
+ "GPU Settings": "Ustawienia GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Ustawia zaawansowane ustawienia GPU, zalecane dla użytkowników z lepszą architekturą GPU.",
+ "GPU Custom Settings": "Niestandardowe ustawienia GPU",
+ "GPU Number": "Numer GPU",
+ "0 to ∞ separated by -": "Od 0 do ∞ oddzielone -",
+ "GPU Information": "Informacje o procesorze GPU",
+ "Pitch Guidance": "Wskazówki dotyczące wysokości dźwięku",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Stosując wskazówki dotyczące wysokości dźwięku, możliwe staje się odzwierciedlenie intonacji oryginalnego głosu, w tym jego wysokości. Ta funkcja jest szczególnie cenna w przypadku śpiewu i innych scenariuszy, w których niezbędne jest zachowanie oryginalnej melodii lub wzoru wysokości dźwięku.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Korzystaj ze wstępnie wytrenowanych modeli podczas trenowania własnych. Takie podejście skraca czas trwania szkolenia i poprawia ogólną jakość.",
+ "Extract Features": "Wyodrębnij funkcje",
+ "Start Training": "Rozpocznij szkolenie",
+ "Generate Index": "Generuj indeks",
+ "Voice Model": "Model głosu",
+ "Select the voice model to use for the conversion.": "Wybierz model głosu, który ma zostać użyty do konwersji.",
+ "Index File": "Plik indeksu",
+ "Select the index file to use for the conversion.": "Wybierz plik indeksu, który ma zostać użyty do konwersji.",
+ "Refresh": "Odświeżyć",
+ "Unload Voice": "Uwolnij głos",
+ "Single": "Pojedynczy",
+ "Upload Audio": "Prześlij dźwięk",
+ "Select Audio": "Wybierz Audio (Dźwięk)",
+ "Select the audio to convert.": "Wybierz dźwięk do konwersji.",
+ "Advanced Settings": "Ustawienia zaawansowane",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Wyczyść wyjścia (usuwa wszystkie pliki audio w zasobach/plikach audio)",
+ "Custom Output Path": "Niestandardowa ścieżka wyjściowa",
+ "Output Path": "Ścieżka wyjściowa",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Ścieżka, w której zostanie zapisany dźwięk wyjściowy, domyślnie w assets/audios/output.wav",
+ "Split Audio": "Podziel dźwięk",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Podziel dźwięk na fragmenty w celu wnioskowania, aby w niektórych przypadkach uzyskać lepsze wyniki.",
+ "Autotune": "Automatyczne dostrajanie",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Zastosuj miękkie autotune do swoich wniosków, zalecane do konwersji śpiewu.",
+ "Clean Audio": "Czysty dźwięk",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Wyczyść wyjście audio za pomocą algorytmów wykrywania szumów, zalecanych do mówienia audio.",
+ "Clean Strength": "Czysta siła",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Ustaw żądany poziom czyszczenia na żądany dźwięk, im bardziej go zwiększysz, tym bardziej się oczyści, ale możliwe, że dźwięk będzie bardziej skompresowany.",
+ "Pitch": "Rzucać",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Ustaw wysokość dźwięku, im wyższa wartość, tym wyższa wysokość.",
+ "Filter Radius": "Promień filtra",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Jeśli liczba ta jest większa lub równa trzem, zastosowanie filtrowania mediany na zebranych wynikach tonów może potencjalnie zmniejszyć oddychanie.",
+ "Search Feature Ratio": "Współczynnik funkcji wyszukiwania",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Wpływ wywierany przez plik indeksu; Wyższa wartość odpowiada większemu wpływowi. Jednak wybranie niższych wartości może pomóc w ograniczeniu artefaktów obecnych w dźwięku.",
+ "Volume Envelope": "Koperta objętości",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Zastąp lub zmieszaj z obwiednią głośności wyjścia. Im współczynnik jest bliższy 1, tym bardziej wykorzystywana jest obwiednia wyjściowa.",
+ "Protect Voiceless Consonants": "Chroń bezdźwięczne spółgłoski",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Chroń wyraźne spółgłoski i dźwięki oddychania, aby zapobiec rozdarciu elektroakustycznemu i innym artefaktom. Pociągnięcie parametru do maksymalnej wartości 0,5 zapewnia kompleksową ochronę. Jednak zmniejszenie tej wartości może zmniejszyć zakres ochrony, jednocześnie potencjalnie łagodząc efekt indeksowania.",
+ "Pitch extraction algorithm": "Algorytm ekstrakcji wysokości dźwięku",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Algorytm ekstrakcji wysokości dźwięku do użycia do konwersji dźwięku. Domyślnym algorytmem jest rmvpe, który jest zalecany w większości przypadków.",
+ "Convert": "Nawrócić",
+ "Export Audio": "Eksportuj dźwięk",
+ "Batch": "Partia",
+ "Input Folder": "Folder wejściowy",
+ "Select the folder containing the audios to convert.": "Wybierz folder zawierający pliki audio do konwersji.",
+ "Enter input path": "Wprowadź ścieżkę wejściową",
+ "Output Folder": "Folder wyjściowy",
+ "Select the folder where the output audios will be saved.": "Wybierz folder, w którym zostaną zapisane wyjściowe pliki audio.",
+ "Enter output path": "Wprowadź ścieżkę wyjściową",
+ "Get information about the audio": "Uzyskiwanie informacji o dźwięku",
+ "Information about the audio file": "Informacje o pliku audio",
+ "Waiting for information...": "Czekam na informację...",
+ "## Voice Blender": "## Blender głosowy",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Wybierz dwa modele brzmienia, ustaw żądany procent mieszania i połącz je w zupełnie nowy głos.",
+ "Voice Blender": "Blender głosu",
+ "Drag and drop your model here": "Przeciągnij i upuść swój model tutaj",
+ "You can also use a custom path.": "Możesz również użyć ścieżki niestandardowej.",
+ "Blend Ratio": "Proporcje mieszania",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Dostosowanie pozycji bardziej w jedną lub drugą stronę sprawi, że model będzie bardziej podobny do pierwszego lub drugiego.",
+ "Fusion": "Fuzja",
+ "Path to Model": "Ścieżka do modelu",
+ "Enter path to model": "Wprowadź ścieżkę do modelu",
+ "Model information to be placed": "Informacje o modelu, które mają zostać umieszczone",
+ "Inroduce the model information": "Zapoznaj się z informacjami o modelu",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Informacje, które mają zostać umieszczone w modelu (możesz pozostawić je puste lub umieścić cokolwiek).",
+ "View model information": "Wyświetlanie informacji o modelu",
+ "Introduce the model pth path": "Wprowadzenie ścieżki pth modelu",
+ "View": "Widok",
+ "Model extraction": "Wyodrębnianie modelu",
+ "Model conversion": "Konwersja modelu",
+ "Pth file": "P-ty plik",
+ "Output of the pth file": "Wyjście pliku pth",
+ "# How to Report an Issue on GitHub": "# Jak zgłosić problem na GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Kliknij przycisk \"Ekran nagrywania\" poniżej, aby rozpocząć nagrywanie napotkanego problemu.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Po zakończeniu nagrywania problemu kliknij przycisk \"Zatrzymaj nagrywanie\" (ten sam przycisk, ale etykieta zmienia się w zależności od tego, czy aktywnie nagrywasz, czy nie).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Przejdź do [GitHub Issues](https://github.com/IAHispano/Applio/issues) i kliknij przycisk \"Nowe zgłoszenie\".",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Wypełnij dostarczony szablon problemu, upewniając się, że w razie potrzeby dołączyłeś szczegóły, i skorzystaj z sekcji zasobów, aby przesłać nagrany plik z poprzedniego kroku.",
+ "Record Screen": "Ekran nagrywania",
+ "Record": "Rekord",
+ "Stop Recording": "Zatrzymaj nagrywanie",
+ "Introduce the model .pth path": "Wprowadzenie ścieżki pth modelu",
+ "See Model Information": "Zobacz informacje o modelu",
+ "## Download Model": "## Pobierz model",
+ "Model Link": "Link do modelu",
+ "Introduce the model link": "Wprowadzenie linku do modelu",
+ "Download Model": "Pobierz model",
+ "## Drop files": "## Upuść pliki",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Przeciągnij plik .pth i plik .index do tego miejsca. Przeciągnij jedną, a potem drugą.",
+ "TTS Voices": "Głosy TTS",
+ "Select the TTS voice to use for the conversion.": "Wybierz głos TTS, który ma być używany do konwersji.",
+ "Text to Synthesize": "Tekst do syntezy",
+ "Enter the text to synthesize.": "Wprowadź tekst do syntezy.",
+ "Or you can upload a .txt file": "Możesz też przesłać plik .txt",
+ "Enter text to synthesize": "Wprowadzanie tekstu do syntezy",
+ "Output Path for TTS Audio": "Ścieżka wyjściowa dla TTS Audio",
+ "Output Path for RVC Audio": "Ścieżka wyjściowa dla dźwięku RVC",
+ "Enable Applio integration with Discord presence": "Włącz integrację Applio z obecnością Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Aktywuje możliwość wyświetlania aktualnej aktywności Applio w Discordzie.",
+ "Enable Applio integration with applio.org/models using flask": "Włączanie integracji aplikacji Applio z applio.org/models przy użyciu kolby",
+ "It will activate the possibility of downloading models with a click from the website.": "Aktywuje możliwość pobierania modeli jednym kliknięciem ze strony internetowej.",
+ "Theme": "Kompozycja",
+ "Select the theme you want to use. (Requires restarting Applio)": "Wybierz motyw, którego chcesz użyć. (Wymaga ponownego uruchomienia aplikacji)",
+ "Language": "Język",
+ "Select the language you want to use. (Requires restarting Applio)": "Wybierz język, którego chcesz używać. (Wymaga ponownego uruchomienia aplikacji)",
+ "Plugin Installer": "Instalator wtyczek",
+ "Drag your plugin.zip to install it": "Przeciągnij plugin.zip, aby go zainstalować",
+ "Version Checker": "Sprawdzanie wersji",
+ "Check which version of Applio is the latest to see if you need to update.": "Sprawdź, która wersja Applio jest najnowsza, aby sprawdzić, czy musisz ją zaktualizować.",
+ "Check for updates": "Sprawdź dostępność aktualizacji"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/pt_BR.json b/assets/i18n/languages/pt_BR.json
new file mode 100644
index 0000000000000000000000000000000000000000..aac7395ad769c25c5c9128b1c639c0bdd71c7e24
--- /dev/null
+++ b/assets/i18n/languages/pt_BR.json
@@ -0,0 +1,113 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "A melhor ferramenta de clonagem de voz, meticulosamente otimizada para potência incomparável, modularidade e experiência amigável.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Esta seção contém alguns utilitários extras que muitas vezes podem estar em fases experimentais.",
+ "Output Information": "Informações de saída",
+ "Inference": "Inferência",
+ "Train": "Treinar",
+ "Extra": "Extra",
+ "Merge Audios": "Mesclar áudios",
+ "Processing": "Processamento",
+ "Audio Analyzer": "Analisador de áudio",
+ "Model Information": "Informação do modelo",
+ "Download": "Baixar",
+ "Report a Bug": "Reportar um Bug",
+ "Preprocess": "Pré-processo",
+ "Model Name": "Nome do modelo",
+ "Enter model name": "Insira o nome do modelo",
+ "Dataset Path": "Caminho do dataset",
+ "Enter dataset path": "Insira o caminho do dataset",
+ "Sampling Rate": "Taxa de amostragem",
+ "RVC Version": "Versão RVC",
+ "Preprocess Dataset": "Pré-processar dataset",
+ "Extract": "Extrair",
+ "Hop Length": "Comprimento do Hop",
+ "Batch Size": "Tamanho do lote",
+ "Save Every Epoch": "Salve Cada Epoch",
+ "Total Epoch": "Epoch Total",
+ "Pretrained": "Pré-treinamento",
+ "Save Only Latest": "Salvar Apenas o último",
+ "Save Every Weights": "Salvar todos os Weights",
+ "Custom Pretrained": "Pretrain personalizado",
+ "Upload Pretrained Model": "Carregar Pretrain",
+ "Pretrained Custom Settings": "Configurações personalizadas do pretrain",
+ "The file you dropped is not a valid pretrained file. Please try again.": "O arquivo que você soltou não é um arquivo de pretrain válido. Por favor, tente novamente.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Clique no botão Atualizar para ver o arquivo pretrain no menu suspenso.",
+ "Pretrained G Path": "Personalizado Pré-treinado G",
+ "Pretrained D Path": "Personalizado Pré-treinado D",
+ "GPU Settings": "Configurações da GPU",
+ "GPU Custom Settings": "Configurações personalizadas da GPU",
+ "GPU Number": "Número da GPU",
+ "0 to ∞ separated by -": "0 a ∞ separados por -",
+ "GPU Information": "Informações da GPU",
+ "Pitch Guidance": "Orientação de Pitch",
+ "Extract Features": "Extrair recursos",
+ "Start Training": "Iniciar Treinamento",
+ "Generate Index": "Gerar Index",
+ "Voice Model": "Modelo de voz",
+ "Index File": "Arquivo de Index",
+ "Refresh": "Atualizar",
+ "Unload Voice": "Descarregar voz",
+ "Single": "Único",
+ "Upload Audio": "Carregar áudio",
+ "Select Audio": "Selecione Áudio",
+ "Advanced Settings": "Configurações avançadas",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Limpar saídas (exclui todos os áudios em ativos/áudios)",
+ "Custom Output Path": "Caminho de saída personalizado",
+ "Output Path": "Caminho de saída",
+ "Pitch": "Pitch",
+ "If >=3: apply median filtering to the harvested pitch results. The value represents the filter radius and can reduce breathiness": "Se >=3: aplicar filtragem mediana aos resultados do pitch colhido. O valor representa o raio do filtro e pode reduzir a soprosidade",
+ "Search Feature Ratio": "Proporção de recursos de Index",
+ "Pitch extraction algorithm": "Algoritmo de extração de pitch",
+ "Convert": "Converter",
+ "Export Audio": "Exportar áudio",
+ "Batch": "Lote",
+ "Input Folder": "Pasta de entrada",
+ "Enter input path": "Insira o caminho de entrada",
+ "Output Folder": "Pasta de saída",
+ "Enter output path": "Insira o caminho de saída",
+ "Get information about the audio": "Obter informações sobre o áudio",
+ "Information about the audio file": "Informações sobre o arquivo de áudio",
+ "Waiting for information...": "À espera de informações...",
+ "Model fusion": "Fusão de modelos",
+ "Weight for Model A": "Peso para o modelo A",
+ "Whether the model has pitch guidance": "Se o modelo tem orientação de pitch",
+ "Model architecture version": "Versão da arquitetura do modelo",
+ "Path to Model A": "Caminho para o Modelo A",
+ "Path to Model B": "Caminho para o Modelo B",
+ "Path to model": "Caminho para o modelo",
+ "Model information to be placed": "Modelo de informação a colocar",
+ "Fusion": "Fusão",
+ "Modify model information": "Modificar informações do modelo",
+ "Path to Model": "Caminho para o modelo",
+ "Model information to be modified": "Modelo de informação a modificar",
+ "Save file name": "Guardar nome de ficheiro",
+ "Modify": "Modificar",
+ "View model information": "Ver informações do modelo",
+ "View": "View",
+ "Model extraction": "Extração do modelo",
+ "Model conversion": "Conversão de modelo",
+ "Pth file": "Arquivo Pth",
+ "Output of the pth file": "Saída do arquivo pth",
+ "# How to Report an Issue on GitHub": "# Como relatar um problema no GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Clique no botão 'Gravar tela' abaixo para começar a gravar o problema que você está enfrentando.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Depois de terminar de gravar o problema, clique no botão 'Parar gravação' (o mesmo botão, mas a etiqueta muda dependendo se você está gravando ativamente ou não).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Vá para [GitHub Issues](https://github.com/IAHispano/Applio/issues) e clique no botão 'New Issue'.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Preencha o modelo de problema fornecido, garantindo incluir detalhes conforme necessário, e utilize a seção de ativos para carregar o arquivo gravado da etapa anterior.",
+ "Record Screen": "Gravar tela",
+ "Record": "Gravar",
+ "Stop Recording": "Parar gravação",
+ "Introduce the model .pth path": "Introduza o caminho .pth do modelo",
+ "See Model Information": "Consulte as informações do modelo",
+ "## Download Model": "## Baixar Modelo",
+ "Model Link": "Link do modelo",
+ "Introduce the model link": "Introduza o link do modelo",
+ "Download Model": "Baixar Modelo",
+ "## Drop files": "## Soltar arquivos",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Arraste o arquivo .pth e o arquivo .index para este espaço. Arraste um e depois o outro.",
+ "TTS Voices": "Vozes TTS",
+ "Text to Synthesize": "Texto a sintetizar",
+ "Enter text to synthesize": "Insira texto para sintetizar",
+ "Output Path for TTS Audio": "Caminho de saída para áudio TTS",
+ "Output Path for RVC Audio": "Caminho de saída para áudio RVC",
+ "Enable Applio integration with Discord presence": "Presença do Applio"
+}
diff --git a/assets/i18n/languages/pt_PT.json b/assets/i18n/languages/pt_PT.json
new file mode 100644
index 0000000000000000000000000000000000000000..487bd045c960825ccd47ca597731118fc998d3ca
--- /dev/null
+++ b/assets/i18n/languages/pt_PT.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "A melhor ferramenta de clonagem de voz, meticulosamente otimizada para potência, modularidade e experiência de fácil utilização incomparáveis.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Esta seção contém alguns utilitários extras que muitas vezes podem estar em fases experimentais.",
+ "Output Information": "Informações de saída",
+ "The output information will be displayed here.": "As informações de saída serão exibidas aqui.",
+ "Inference": "Inferência",
+ "Train": "Trem",
+ "Extra": "Extra",
+ "Merge Audios": "Mesclar áudios",
+ "Processing": "Processamento",
+ "Audio Analyzer": "Analisador de áudio",
+ "Model Information": "Informações do modelo",
+ "Plugins": "Plug-ins",
+ "Download": "Baixar",
+ "Report a Bug": "Relatar um bug",
+ "Settings": "Configurações",
+ "Preprocess": "Pré-processar",
+ "Model Name": "Nome do modelo",
+ "Name of the new model.": "Nome do novo modelo.",
+ "Enter model name": "Digite o nome do modelo",
+ "Dataset Path": "Caminho do conjunto de dados",
+ "Path to the dataset folder.": "Caminho para a pasta do conjunto de dados.",
+ "Refresh Datasets": "Atualizar conjuntos de dados",
+ "Dataset Creator": "Criador de conjunto de dados",
+ "Dataset Name": "Nome do conjunto de dados",
+ "Name of the new dataset.": "Nome do novo conjunto de dados.",
+ "Enter dataset name": "Insira o nome do conjunto de dados",
+ "Upload Audio Dataset": "Carregar conjunto de dados de áudio",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "O arquivo de áudio foi adicionado com êxito ao conjunto de dados. Clique no botão de pré-processo.",
+ "Enter dataset path": "Inserir caminho do conjunto de dados",
+ "Sampling Rate": "Taxa de amostragem",
+ "The sampling rate of the audio files.": "A taxa de amostragem dos arquivos de áudio.",
+ "RVC Version": "Versão RVC",
+ "The RVC version of the model.": "A versão RVC do modelo.",
+ "Preprocess Dataset": "Pré-processar conjunto de dados",
+ "Extract": "Extrair",
+ "Hop Length": "Comprimento do salto",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Denota a duração necessária para que o sistema faça a transição para uma mudança de tom significativa. Comprimentos de salto menores requerem mais tempo para inferência, mas tendem a produzir maior precisão de pitch.",
+ "Batch Size": "Tamanho do lote",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "É aconselhável alinhá-lo com a VRAM disponível da sua GPU. Uma configuração de 4 oferece precisão aprimorada, mas processamento mais lento, enquanto 8 fornece resultados mais rápidos e padrão.",
+ "Save Every Epoch": "Salve todas as épocas",
+ "Determine at how many epochs the model will saved at.": "Determine em quantas épocas o modelo será salvo.",
+ "Total Epoch": "Época Total",
+ "Specifies the overall quantity of epochs for the model training process.": "Especifica a quantidade total de épocas para o processo de treinamento do modelo.",
+ "Pretrained": "Pré-treinado",
+ "Save Only Latest": "Salvar somente as últimas",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Habilitar essa configuração resultará nos arquivos G e D salvando apenas suas versões mais recentes, efetivamente conservando espaço de armazenamento.",
+ "Save Every Weights": "Economize todos os pesos",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Essa configuração permite que você salve os pesos do modelo na conclusão de cada época.",
+ "Custom Pretrained": "Pré-treinado personalizado",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "A utilização de modelos pré-treinados personalizados pode levar a resultados superiores, pois a seleção dos modelos pré-treinados mais adequados adaptados ao caso de uso específico pode melhorar significativamente o desempenho.",
+ "Upload Pretrained Model": "Carregar modelo pré-treinado",
+ "Refresh Custom Pretraineds": "Atualizar pré-treinados personalizados",
+ "Pretrained Custom Settings": "Configurações personalizadas pré-treinadas",
+ "The file you dropped is not a valid pretrained file. Please try again.": "O arquivo descartado não é um arquivo pré-treinado válido. Tente novamente.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Clique no botão Atualizar para ver o arquivo pré-treinado no menu suspenso.",
+ "Pretrained G Path": "G pré-treinado personalizado",
+ "Pretrained D Path": "D pré-treinado personalizado",
+ "GPU Settings": "Configurações da GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Define configurações avançadas de GPU, recomendadas para usuários com melhor arquitetura de GPU.",
+ "GPU Custom Settings": "Configurações personalizadas da GPU",
+ "GPU Number": "Número da GPU",
+ "0 to ∞ separated by -": "0 a ∞ separados por -",
+ "GPU Information": "Informações da GPU",
+ "Pitch Guidance": "Orientação de Pitch",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Ao empregar a orientação de pitch, torna-se viável espelhar a entonação da voz original, incluindo seu pitch. Esta característica é particularmente valiosa para o canto e outros cenários onde preservar a melodia original ou padrão de tom é essencial.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Utilize modelos pré-treinados ao treinar seus próprios. Essa abordagem reduz a duração do treinamento e melhora a qualidade geral.",
+ "Extract Features": "Recursos de extração",
+ "Start Training": "Comece a Treinar",
+ "Generate Index": "Gerar índice",
+ "Voice Model": "Modelo de Voz",
+ "Select the voice model to use for the conversion.": "Selecione o modelo de voz a ser usado para a conversão.",
+ "Index File": "Arquivo de índice",
+ "Select the index file to use for the conversion.": "Selecione o arquivo de índice a ser usado para a conversão.",
+ "Refresh": "Atualizar",
+ "Unload Voice": "Descarregar voz",
+ "Single": "Único",
+ "Upload Audio": "Carregar áudio",
+ "Select Audio": "Selecione Áudio",
+ "Select the audio to convert.": "Selecione o áudio a ser convertido.",
+ "Advanced Settings": "Configurações avançadas",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Limpar saídas (exclui todos os áudios em ativos/áudios)",
+ "Custom Output Path": "Caminho de saída personalizado",
+ "Output Path": "Caminho de saída",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "O caminho onde o áudio de saída será salvo, por padrão em ativos/áudios/output.wav",
+ "Split Audio": "Áudio dividido",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Divida o áudio em pedaços para inferência para obter melhores resultados em alguns casos.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Aplique um autotune suave às suas inferências, recomendado para conversões de canto.",
+ "Clean Audio": "Áudio limpo",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Limpe sua saída de áudio usando algoritmos de detecção de ruído, recomendados para falar áudios.",
+ "Clean Strength": "Força Limpa",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Defina o nível de limpeza para o áudio desejado, quanto mais você aumentá-lo, mais ele será limpo, mas é possível que o áudio seja mais compactado.",
+ "Pitch": "Campo",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Defina o tom do áudio, quanto maior o valor, maior o pitch.",
+ "Filter Radius": "Raio do filtro",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Se o número for maior ou igual a três, empregar a filtragem mediana nos resultados de tom coletados tem o potencial de diminuir a respiração.",
+ "Search Feature Ratio": "Taxa de recursos de pesquisa",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Influência exercida pelo arquivo de índice; quanto maior o valor corresponde maior a influência. No entanto, optar por valores mais baixos pode ajudar a mitigar artefatos presentes no áudio.",
+ "Volume Envelope": "Volume Envelope",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Substitua ou misture com o envelope de volume da saída. Quanto mais próxima a proporção estiver de 1, mais o envelope de saída será empregado.",
+ "Protect Voiceless Consonants": "Proteja as consoantes surdas",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Proteja consoantes distintas e sons respiratórios para evitar rasgos eletroacústicos e outros artefatos. Puxar o parâmetro para seu valor máximo de 0,5 oferece proteção abrangente. No entanto, a redução desse valor pode diminuir a extensão da proteção e, ao mesmo tempo, potencialmente atenuar o efeito de indexação.",
+ "Pitch extraction algorithm": "Algoritmo de extração de pitch",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Algoritmo de extração de pitch para usar na conversão de áudio. O algoritmo padrão é rmvpe, que é recomendado para a maioria dos casos.",
+ "Convert": "Converter",
+ "Export Audio": "Exportar áudio",
+ "Batch": "Lote",
+ "Input Folder": "Pasta de entrada",
+ "Select the folder containing the audios to convert.": "Selecione a pasta que contém os áudios a serem convertidos.",
+ "Enter input path": "Insira o caminho de entrada",
+ "Output Folder": "Pasta de saída",
+ "Select the folder where the output audios will be saved.": "Selecione a pasta onde os áudios de saída serão salvos.",
+ "Enter output path": "Insira o caminho de saída",
+ "Get information about the audio": "Obter informações sobre o áudio",
+ "Information about the audio file": "Informações sobre o arquivo de áudio",
+ "Waiting for information...": "Aguardando informações...",
+ "## Voice Blender": "## Liquidificador de Voz",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Selecione dois modelos de voz, defina a porcentagem de mistura desejada e misture-os em uma voz totalmente nova.",
+ "Voice Blender": "Liquidificador de voz",
+ "Drag and drop your model here": "Arraste e solte seu modelo aqui",
+ "You can also use a custom path.": "Você também pode usar um caminho personalizado.",
+ "Blend Ratio": "Proporção de mistura",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Ajustar a posição mais para um lado ou para o outro tornará o modelo mais semelhante ao primeiro ou ao segundo.",
+ "Fusion": "Fusão",
+ "Path to Model": "Caminho para o modelo",
+ "Enter path to model": "Insira o caminho para o modelo",
+ "Model information to be placed": "Informações do modelo a ser colocado",
+ "Inroduce the model information": "Inroduce as informações do modelo",
+ "The information to be placed in the model (You can leave it blank or put anything).": "As informações a serem colocadas no modelo (Você pode deixá-lo em branco ou colocar qualquer coisa).",
+ "View model information": "Exibir informações do modelo",
+ "Introduce the model pth path": "Apresentar o caminho pth do modelo",
+ "View": "Vista",
+ "Model extraction": "Extração do modelo",
+ "Model conversion": "Conversão de modelos",
+ "Pth file": "Arquivo Pth",
+ "Output of the pth file": "Saída do arquivo pth",
+ "# How to Report an Issue on GitHub": "# Como relatar um problema no GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Clique no botão 'Gravar tela' abaixo para começar a gravar o problema que você está enfrentando.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Depois de terminar de gravar o problema, clique no botão 'Parar gravação' (o mesmo botão, mas a etiqueta muda dependendo se você está gravando ativamente ou não).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Vá para [Problemas do GitHub](https://github.com/IAHispano/Applio/issues) e clique no botão 'Novo problema'.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Preencha o modelo de problema fornecido, garantindo incluir detalhes conforme necessário, e utilize a seção de ativos para carregar o arquivo gravado da etapa anterior.",
+ "Record Screen": "Tela de gravação",
+ "Record": "Registro",
+ "Stop Recording": "Parar gravação",
+ "Introduce the model .pth path": "Apresentar o caminho .pth do modelo",
+ "See Model Information": "Ver informações do modelo",
+ "## Download Model": "## Baixar Modelo",
+ "Model Link": "Link do modelo",
+ "Introduce the model link": "Apresentar o link do modelo",
+ "Download Model": "Download do Modelo",
+ "## Drop files": "## Soltar arquivos",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Arraste o arquivo .pth e o arquivo .index para este espaço. Arraste um e depois o outro.",
+ "TTS Voices": "Vozes TTS",
+ "Select the TTS voice to use for the conversion.": "Selecione a voz TTS a ser usada para a conversão.",
+ "Text to Synthesize": "Texto para sintetizar",
+ "Enter the text to synthesize.": "Digite o texto a ser sintetizado.",
+ "Or you can upload a .txt file": "Ou você pode carregar um arquivo .txt",
+ "Enter text to synthesize": "Digite o texto para sintetizar",
+ "Output Path for TTS Audio": "Caminho de saída para áudio TTS",
+ "Output Path for RVC Audio": "Caminho de saída para áudio RVC",
+ "Enable Applio integration with Discord presence": "Habilitar a integração do Applio com a presença do Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Ele ativará a possibilidade de exibir a atividade atual do Applio no Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Habilitar a integração do Applio com applio.org/models usando frasco",
+ "It will activate the possibility of downloading models with a click from the website.": "Ele ativará a possibilidade de baixar modelos com um clique no site.",
+ "Theme": "Tema",
+ "Select the theme you want to use. (Requires restarting Applio)": "Selecione o tema que deseja usar. (Requer reiniciar o Applio)",
+ "Language": "Idioma",
+ "Select the language you want to use. (Requires restarting Applio)": "Selecione o idioma que deseja usar. (Requer reiniciar o Applio)",
+ "Plugin Installer": "Instalador de Plug-ins",
+ "Drag your plugin.zip to install it": "Arraste o plugin.zip para instalá-lo",
+ "Version Checker": "Verificador de versão",
+ "Check which version of Applio is the latest to see if you need to update.": "Verifique qual versão do Applio é a mais recente para ver se você precisa atualizar.",
+ "Check for updates": "Verificar se há atualizações"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/ro_RO.json b/assets/i18n/languages/ro_RO.json
new file mode 100644
index 0000000000000000000000000000000000000000..473bcf9d5da82790fa50eb5aa8770eb13c92b2e4
--- /dev/null
+++ b/assets/i18n/languages/ro_RO.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Cel mai bun instrument de clonare a vocii, optimizat meticulos pentru putere, modularitate și experiență ușor de utilizat.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Această secțiune conține câteva utilități suplimentare care pot fi adesea în faze experimentale.",
+ "Output Information": "Informații despre ieșire",
+ "The output information will be displayed here.": "Informațiile de ieșire vor fi afișate aici.",
+ "Inference": "Deducţie",
+ "Train": "Tren",
+ "Extra": "Superfluu",
+ "Merge Audios": "Îmbinare audio",
+ "Processing": "Prelucrare",
+ "Audio Analyzer": "Analizor audio",
+ "Model Information": "Informații despre model",
+ "Plugins": "Plugin-uri",
+ "Download": "Descărca",
+ "Report a Bug": "Raportați o eroare",
+ "Settings": "Setări",
+ "Preprocess": "Preproces",
+ "Model Name": "Numele modelului",
+ "Name of the new model.": "Numele noului model.",
+ "Enter model name": "Introduceți numele modelului",
+ "Dataset Path": "Calea setului de date",
+ "Path to the dataset folder.": "Calea către folderul setului de date.",
+ "Refresh Datasets": "Reîmprospătarea seturilor de date",
+ "Dataset Creator": "Creator de seturi de date",
+ "Dataset Name": "Numele setului de date",
+ "Name of the new dataset.": "Numele noului set de date.",
+ "Enter dataset name": "Introduceți numele setului de date",
+ "Upload Audio Dataset": "Încărcați setul de date audio",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Fișierul audio a fost adăugat cu succes la setul de date. Vă rugăm să faceți clic pe butonul de preprocesare.",
+ "Enter dataset path": "Introduceți calea setului de date",
+ "Sampling Rate": "Rata de eșantionare",
+ "The sampling rate of the audio files.": "Rata de eșantionare a fișierelor audio.",
+ "RVC Version": "Versiunea RVC",
+ "The RVC version of the model.": "Versiunea RVC a modelului.",
+ "Preprocess Dataset": "Set de date preproces",
+ "Extract": "Extract",
+ "Hop Length": "Lungimea hameiului",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Denotă durata necesară pentru ca sistemul să treacă la o schimbare semnificativă a înălțimii. Lungimile mai mici ale hameiului necesită mai mult timp pentru inferență, dar tind să producă o precizie mai mare a pasului.",
+ "Batch Size": "Mărimea lotului",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Este recomandabil să îl aliniați cu VRAM disponibil al GPU-ului. O setare de 4 oferă o precizie îmbunătățită, dar o procesare mai lentă, în timp ce 8 oferă rezultate mai rapide și standard.",
+ "Save Every Epoch": "Salvați fiecare epocă",
+ "Determine at how many epochs the model will saved at.": "Determinați la câte epoci va fi salvat modelul.",
+ "Total Epoch": "Epoca totală",
+ "Specifies the overall quantity of epochs for the model training process.": "Specifică numărul total de epoci pentru procesul de instruire a modelului.",
+ "Pretrained": "Preinstruit",
+ "Save Only Latest": "Salvați numai cele mai recente",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Activarea acestei setări va avea ca rezultat salvarea fișierelor G și D numai a celor mai recente versiuni, economisind în mod eficient spațiul de stocare.",
+ "Save Every Weights": "Economisiți fiecare greutate",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Această setare vă permite să economisiți greutățile modelului la sfârșitul fiecărei epoci.",
+ "Custom Pretrained": "Personalizat Pretrained",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Utilizarea modelelor personalizate pre-antrenate poate duce la rezultate superioare, deoarece selectarea celor mai potrivite modele pre-antrenate adaptate cazului de utilizare specific poate îmbunătăți semnificativ performanța.",
+ "Upload Pretrained Model": "Încărcați modelul preinstruit",
+ "Refresh Custom Pretraineds": "Reîmprospătați preinstruiții personalizați",
+ "Pretrained Custom Settings": "Setări personalizate pre-instruite",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Fișierul pe care l-ați scăpat nu este un fișier preinstruit valid. Vă rugăm să încercați din nou.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Faceți clic pe butonul de reîmprospătare pentru a vedea fișierul preantrenat în meniul derulant.",
+ "Pretrained G Path": "G personalizat preantrenat",
+ "Pretrained D Path": "Personalizat Pretrained D",
+ "GPU Settings": "Setări GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Setează setări GPU avansate, recomandate utilizatorilor cu o arhitectură GPU mai bună.",
+ "GPU Custom Settings": "Setări personalizate GPU",
+ "GPU Number": "Număr GPU",
+ "0 to ∞ separated by -": "0 până la ∞ separate de -",
+ "GPU Information": "Informații GPU",
+ "Pitch Guidance": "Pitch Guidance",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Prin utilizarea ghidării tonului, devine fezabilă oglindirea intonației vocii originale, inclusiv a înălțimii acesteia. Această caracteristică este deosebit de valoroasă pentru cântat și alte scenarii în care păstrarea melodiei originale sau a modelului de înălțime este esențială.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Utilizați modele pre-antrenate atunci când vă antrenați propriul. Această abordare reduce durata antrenamentului și îmbunătățește calitatea generală.",
+ "Extract Features": "Extrageți caracteristicile",
+ "Start Training": "Începeți instruirea",
+ "Generate Index": "Generare index",
+ "Voice Model": "Model vocal",
+ "Select the voice model to use for the conversion.": "Selectați modelul vocal de utilizat pentru conversie.",
+ "Index File": "Fișier index",
+ "Select the index file to use for the conversion.": "Selectați fișierul index de utilizat pentru conversie.",
+ "Refresh": "Împrospăta",
+ "Unload Voice": "Descărcare voce",
+ "Single": "Singur",
+ "Upload Audio": "Încărcare audio",
+ "Select Audio": "Selectați Audio",
+ "Select the audio to convert.": "Selectați sunetul de convertit.",
+ "Advanced Settings": "Setări avansate",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Ștergeți ieșirile (Șterge toate audio-urile din active / audio)",
+ "Custom Output Path": "Cale de ieșire personalizată",
+ "Output Path": "Cale de ieșire",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Calea în care va fi salvat sunetul de ieșire, în mod implicit în active / audio / output.wav",
+ "Split Audio": "Împărțire audio",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Împărțiți sunetul în bucăți pentru inferență pentru a obține rezultate mai bune în unele cazuri.",
+ "Autotune": "Reglare automată",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Aplicați o reglare automată ușoară la inferențele dvs., recomandată pentru conversiile de cântat.",
+ "Clean Audio": "Sunet curat",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Curățați ieșirea audio utilizând algoritmi de detectare a zgomotului, recomandați pentru enunțarea sunetului.",
+ "Clean Strength": "Rezistență curată",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Setați nivelul de curățare la sunetul dorit, cu cât îl măriți mai mult, cu atât se va curăța mai mult, dar este posibil ca sunetul să fie mai comprimat.",
+ "Pitch": "Smoală",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Setați înălțimea sunetului, cu cât este mai mare valoarea, cu atât este mai mare înălțimea.",
+ "Filter Radius": "Raza filtrului",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Dacă numărul este mai mare sau egal cu trei, utilizarea filtrării mediane pe rezultatele tonului colectat are potențialul de a scădea respirația.",
+ "Search Feature Ratio": "Raportul caracteristicilor de căutare",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Influența exercitată de fișierul index; O valoare mai mare corespunde unei influențe mai mari. Cu toate acestea, optarea pentru valori mai mici poate ajuta la atenuarea artefactelor prezente în audio.",
+ "Volume Envelope": "Plic de volum",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Înlocuiți sau amestecați cu plicul de volum al ieșirii. Cu cât raportul este mai apropiat de 1, cu atât este folosit mai mult plicul de ieșire.",
+ "Protect Voiceless Consonants": "Protejați consoanele fără voce",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Protejați consoanele distincte și sunetele de respirație pentru a preveni ruperea electro-acustică și alte artefacte. Tragerea parametrului la valoarea maximă de 0,5 oferă o protecție completă. Cu toate acestea, reducerea acestei valori ar putea reduce gradul de protecție, atenuând în același timp efectul de indexare.",
+ "Pitch extraction algorithm": "Algoritm de extracție a pitch-ului",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Algoritm de extragere a pitch-ului de utilizat pentru conversia audio. Algoritmul implicit este rmvpe, care este recomandat în majoritatea cazurilor.",
+ "Convert": "Converti",
+ "Export Audio": "Exportați audio",
+ "Batch": "Lot",
+ "Input Folder": "Folder de intrare",
+ "Select the folder containing the audios to convert.": "Selectați folderul care conține audio-urile de convertit.",
+ "Enter input path": "Introduceți calea de intrare",
+ "Output Folder": "Dosar de ieșire",
+ "Select the folder where the output audios will be saved.": "Selectați folderul în care vor fi salvate audio-urile de ieșire.",
+ "Enter output path": "Introduceți calea de ieșire",
+ "Get information about the audio": "Obțineți informații despre audio",
+ "Information about the audio file": "Informații despre fișierul audio",
+ "Waiting for information...": "În așteptarea informațiilor...",
+ "## Voice Blender": "## Blender de voce",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Selectați două modele de voce, setați procentajul de amestec dorit și amestecați-le într-o voce complet nouă.",
+ "Voice Blender": "Blender de voce",
+ "Drag and drop your model here": "Glisați și fixați modelul aici",
+ "You can also use a custom path.": "De asemenea, puteți utiliza un traseu personalizat.",
+ "Blend Ratio": "Raport de amestecare",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Ajustarea poziției mai mult spre o parte sau alta va face modelul mai asemănător cu primul sau al doilea.",
+ "Fusion": "Fuziune",
+ "Path to Model": "Calea către model",
+ "Enter path to model": "Introduceți calea către model",
+ "Model information to be placed": "Informații despre model care trebuie plasate",
+ "Inroduce the model information": "Inroduceți informațiile despre model",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Informațiile care trebuie plasate în model (Puteți să-l lăsați necompletat sau să puneți orice).",
+ "View model information": "Vizualizarea informațiilor despre model",
+ "Introduce the model pth path": "Introduceți calea pth a modelului",
+ "View": "Vedere",
+ "Model extraction": "Extragerea modelului",
+ "Model conversion": "Conversia modelului",
+ "Pth file": "Fișier Pth",
+ "Output of the pth file": "Ieșirea fișierului pth",
+ "# How to Report an Issue on GitHub": "# Cum să raportați o problemă pe GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Faceți clic pe butonul \"Ecran de înregistrare\" de mai jos pentru a începe înregistrarea problemei pe care o întâmpinați.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. După ce ați terminat de înregistrat problema, faceți clic pe butonul \"Opriți înregistrarea\" (același buton, dar eticheta se schimbă în funcție de înregistrarea activă sau nu).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Accesați [Probleme GitHub] (https://github.com/IAHispano/Applio/issues) și faceți clic pe butonul \"Problemă nouă\".",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Completați șablonul de problemă furnizat, asigurându-vă că includeți detalii după cum este necesar și utilizați secțiunea active pentru a încărca fișierul înregistrat din pasul anterior.",
+ "Record Screen": "Ecran de înregistrare",
+ "Record": "Disc",
+ "Stop Recording": "Opriți înregistrarea",
+ "Introduce the model .pth path": "Introducerea căii .pth a modelului",
+ "See Model Information": "Consultați informațiile despre model",
+ "## Download Model": "## Descărcați modelul",
+ "Model Link": "Model Link",
+ "Introduce the model link": "Introduceți linkul modelului",
+ "Download Model": "Descărcați modelul",
+ "## Drop files": "## Aruncați fișiere",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Glisați fișierul .pth și fișierul .index în acest spațiu. Trageți unul și apoi celălalt.",
+ "TTS Voices": "Voci TTS",
+ "Select the TTS voice to use for the conversion.": "Selectați vocea TTS de utilizat pentru conversie.",
+ "Text to Synthesize": "Text pentru sintetizare",
+ "Enter the text to synthesize.": "Introduceți textul pentru a sintetiza.",
+ "Or you can upload a .txt file": "Sau puteți încărca un fișier .txt",
+ "Enter text to synthesize": "Introduceți text pentru sintetizare",
+ "Output Path for TTS Audio": "Cale de ieșire pentru TTS Audio",
+ "Output Path for RVC Audio": "Cale de ieșire pentru RVC Audio",
+ "Enable Applio integration with Discord presence": "Activați integrarea Applio cu prezența Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Acesta va activa posibilitatea afișării activității curente Applio în Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Activați integrarea Applio cu applio.org/models folosind balonul",
+ "It will activate the possibility of downloading models with a click from the website.": "Acesta va activa posibilitatea de a descărca modele cu un clic de pe site.",
+ "Theme": "Temă",
+ "Select the theme you want to use. (Requires restarting Applio)": "Selectați tema pe care doriți să o utilizați. (Necesită repornirea Applio)",
+ "Language": "Limbă",
+ "Select the language you want to use. (Requires restarting Applio)": "Selectați limba pe care doriți să o utilizați. (Necesită repornirea Applio)",
+ "Plugin Installer": "Instalator de plugin-uri",
+ "Drag your plugin.zip to install it": "Trageți plugin.zip pentru a-l instala",
+ "Version Checker": "Verificator de versiuni",
+ "Check which version of Applio is the latest to see if you need to update.": "Verificați ce versiune de Applio este cea mai recentă pentru a vedea dacă trebuie să actualizați.",
+ "Check for updates": "Căutați actualizări"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/ru_RU.json b/assets/i18n/languages/ru_RU.json
new file mode 100644
index 0000000000000000000000000000000000000000..641127eab84fdd4929afba7865a2905bc505b17c
--- /dev/null
+++ b/assets/i18n/languages/ru_RU.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Идеальный инструмент для клонирования голоса, тщательно оптимизированный для непревзойденной мощности, модульности и удобства использования.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Этот раздел содержит некоторые дополнительные утилиты, которые часто находятся на экспериментальных стадиях.",
+ "Output Information": "Выходная информация",
+ "The output information will be displayed here.": "Здесь будет отображена выходная информация.",
+ "Inference": "Вывод",
+ "Train": "Поезд",
+ "Extra": "Дополнительный",
+ "Merge Audios": "Слияние аудиозаписей",
+ "Processing": "Обработка",
+ "Audio Analyzer": "Анализатор звука",
+ "Model Information": "Информация о модели",
+ "Plugins": "Плагины",
+ "Download": "Загружать",
+ "Report a Bug": "Сообщить об ошибке",
+ "Settings": "Параметры",
+ "Preprocess": "Предварительной обработки",
+ "Model Name": "Название модели",
+ "Name of the new model.": "Название новой модели.",
+ "Enter model name": "Введите название модели",
+ "Dataset Path": "Путь к набору данных",
+ "Path to the dataset folder.": "Путь к папке набора данных.",
+ "Refresh Datasets": "Обновление наборов данных",
+ "Dataset Creator": "Создатель набора данных",
+ "Dataset Name": "Имя набора данных",
+ "Name of the new dataset.": "Имя нового набора данных.",
+ "Enter dataset name": "Введите имя набора данных",
+ "Upload Audio Dataset": "Загрузка набора аудиоданных",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Аудиофайл успешно добавлен в набор данных. Пожалуйста, нажмите кнопку предварительной обработки.",
+ "Enter dataset path": "Введите путь к набору данных",
+ "Sampling Rate": "Частота дискретизации",
+ "The sampling rate of the audio files.": "Частота дискретизации аудиофайлов.",
+ "RVC Version": "Версия РВК",
+ "The RVC version of the model.": "Версия модели РВК.",
+ "Preprocess Dataset": "Набор данных предварительной обработки",
+ "Extract": "Экстракт",
+ "Hop Length": "Длина хмеля",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Обозначает время, необходимое системе для перехода к значительному изменению высоты тона. Меньшая длина скачка требует больше времени для логического вывода, но, как правило, обеспечивает более высокую точность шага.",
+ "Batch Size": "Размер партии",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Рекомендуется выровнять его с доступной видеопамятью вашего графического процессора. Значение 4 обеспечивает повышенную точность, но более медленную обработку, в то время как значение 8 обеспечивает более быстрые и стандартные результаты.",
+ "Save Every Epoch": "Сохраняйте каждую эпоху",
+ "Determine at how many epochs the model will saved at.": "Определите, в скольких эпохах будет сохраняться модель.",
+ "Total Epoch": "Общая эпоха",
+ "Specifies the overall quantity of epochs for the model training process.": "Задает общее количество эпох для процесса обучения модели.",
+ "Pretrained": "Предварительно обученный",
+ "Save Only Latest": "Сохранить только последние новости",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Включение этого параметра приведет к тому, что файлы G и D будут сохранять только свои самые последние версии, эффективно экономя место на диске.",
+ "Save Every Weights": "Сохраняйте все веса",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Эта настройка позволяет сохранять весовые коэффициенты модели в конце каждой эпохи.",
+ "Custom Pretrained": "Пользовательский предварительно обученный",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Использование пользовательских предварительно обученных моделей может привести к превосходным результатам, так как выбор наиболее подходящих предварительно обученных моделей, адаптированных к конкретному сценарию использования, может значительно повысить производительность.",
+ "Upload Pretrained Model": "Отправка предварительно обученной модели",
+ "Refresh Custom Pretraineds": "Обновление пользовательских предварительно обученных объектов",
+ "Pretrained Custom Settings": "Предварительно обученные пользовательские параметры",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Файл, который вы удалили, не является допустимым предварительно обученным файлом. Повторите попытку.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Нажмите кнопку обновления, чтобы увидеть предварительно обученный файл в раскрывающемся меню.",
+ "Pretrained G Path": "Пользовательский предварительно обученный G",
+ "Pretrained D Path": "Пользовательский предварительно обученный D",
+ "GPU Settings": "Настройки графического процессора",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Устанавливает расширенные настройки графического процессора, рекомендуемые для пользователей с улучшенной архитектурой графического процессора.",
+ "GPU Custom Settings": "Пользовательские настройки графического процессора",
+ "GPU Number": "Номер графического процессора",
+ "0 to ∞ separated by -": "от 0 до ∞ разделенных -",
+ "GPU Information": "Информация о графическом процессоре",
+ "Pitch Guidance": "Руководство по питчу",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Используя управление высотой тона, становится возможным отразить интонацию исходного голоса, включая его высоту. Эта функция особенно ценна для пения и других сценариев, где важно сохранить оригинальную мелодию или тональность.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Используйте предварительно обученные модели при обучении своих собственных. Такой подход сокращает продолжительность обучения и повышает общее качество.",
+ "Extract Features": "Извлечение объектов",
+ "Start Training": "Начать обучение",
+ "Generate Index": "Сгенерировать индекс",
+ "Voice Model": "Голосовая модель",
+ "Select the voice model to use for the conversion.": "Выберите голосовую модель, которая будет использоваться для преобразования.",
+ "Index File": "Индексный файл",
+ "Select the index file to use for the conversion.": "Выберите индексный файл, который будет использоваться для преобразования.",
+ "Refresh": "Освежать",
+ "Unload Voice": "Выгрузить голос",
+ "Single": "Единственный",
+ "Upload Audio": "Загрузить аудио",
+ "Select Audio": "Выберите Аудио",
+ "Select the audio to convert.": "Выберите аудио для преобразования.",
+ "Advanced Settings": "Расширенные настройки",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Clear Outputs (Удаляет все аудиозаписи в assets/audios)",
+ "Custom Output Path": "Пользовательский выходной путь",
+ "Output Path": "Выходной путь",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Путь, по которому будет сохранен выходной звук, по умолчанию в assets/audios/output.wav",
+ "Split Audio": "Разделенное аудио",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Разделите аудио на фрагменты для вывода, чтобы получить лучшие результаты в некоторых случаях.",
+ "Autotune": "Автотюн",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Примените мягкую автонастройку к своим выводам, рекомендуемую для певческих преобразований.",
+ "Clean Audio": "Чистый звук",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Очистите аудиовыход с помощью алгоритмов обнаружения шума, рекомендуемых для проговаривания аудио.",
+ "Clean Strength": "Чистая прочность",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Установите желаемый уровень очистки звука, чем больше вы его увеличите, тем больше он будет очищаться, но возможно, что звук будет более сжатым.",
+ "Pitch": "Смола",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Установите высоту звука, чем выше значение, тем выше высота тона.",
+ "Filter Radius": "Радиус фильтра",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Если это число больше или равно трем, использование медианной фильтрации по собранным результатам тона может привести к снижению дыхания.",
+ "Search Feature Ratio": "Соотношение объектов поиска",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Влияние, оказываемое индексным файлом; Чем выше значение, тем больше влияние. Однако выбор более низких значений может помочь смягчить артефакты, присутствующие в аудио.",
+ "Volume Envelope": "Огибающая объема",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Замените или смешайте с огибающей громкости выхода. Чем ближе отношение к 1, тем больше используется выходная огибающая.",
+ "Protect Voiceless Consonants": "Защита глухих согласных",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Защитите отчетливые согласные и звуки дыхания, чтобы предотвратить электроакустические разрывы и другие артефакты. Извлечение параметра до максимального значения 0,5 обеспечивает комплексную защиту. Однако уменьшение этого значения может снизить степень защиты, потенциально смягчив эффект индексирования.",
+ "Pitch extraction algorithm": "Алгоритм извлечения высоты тона",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Алгоритм извлечения высоты тона, используемый для преобразования звука. По умолчанию используется алгоритм rmvpe, который рекомендуется для большинства случаев.",
+ "Convert": "Обращать",
+ "Export Audio": "Экспорт аудио",
+ "Batch": "Партия",
+ "Input Folder": "Входная папка",
+ "Select the folder containing the audios to convert.": "Выберите папку, содержащую аудиофайлы для преобразования.",
+ "Enter input path": "Введите путь ввода",
+ "Output Folder": "Выходная папка",
+ "Select the folder where the output audios will be saved.": "Выберите папку, в которой будут сохранены выходные аудиозаписи.",
+ "Enter output path": "Введите выходной путь",
+ "Get information about the audio": "Получение информации об аудио",
+ "Information about the audio file": "Информация об аудиофайле",
+ "Waiting for information...": "Жду информации...",
+ "## Voice Blender": "## Голосовой блендер",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Выберите две модели голоса, установите желаемый процент смешивания и смешайте их в совершенно новый голос.",
+ "Voice Blender": "Голосовой блендер",
+ "Drag and drop your model here": "Перетащите сюда свою модель",
+ "You can also use a custom path.": "Вы также можете использовать пользовательский путь.",
+ "Blend Ratio": "Соотношение смешивания",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Изменение положения в ту или иную сторону сделает модель более похожей на первую или вторую.",
+ "Fusion": "Слияние",
+ "Path to Model": "Путь к модели",
+ "Enter path to model": "Введите путь к модели",
+ "Model information to be placed": "Информация о модели, которая будет размещена",
+ "Inroduce the model information": "Ввод информации о модели",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Информация, которая будет размещена в модели (Вы можете оставить ее пустой или поставить что угодно).",
+ "View model information": "Просмотр информации о модели",
+ "Introduce the model pth path": "Знакомство с моделью pth-пути",
+ "View": "Вид",
+ "Model extraction": "Извлечение модели",
+ "Model conversion": "Преобразование модели",
+ "Pth file": "P-й файл",
+ "Output of the pth file": "Вывод p-го файла",
+ "# How to Report an Issue on GitHub": "# Как сообщить о проблеме на GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Нажмите кнопку «Записать экран» ниже, чтобы начать запись проблемы, с которой вы столкнулись.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. После того, как вы закончили запись задачи, нажмите кнопку «Остановить запись» (та же кнопка, но метка меняется в зависимости от того, ведете ли вы активную запись или нет).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Перейдите в [GitHub Issues](https://github.com/IAHispano/Applio/issues) и нажмите кнопку «Новая проблема».",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Заполните предоставленный шаблон задачи, не забудьте включить необходимые сведения и используйте раздел ресурсов для загрузки записанного файла с предыдущего шага.",
+ "Record Screen": "Запись экрана",
+ "Record": "Запись",
+ "Stop Recording": "Остановить запись",
+ "Introduce the model .pth path": "Знакомство с моделью .pth-пути",
+ "See Model Information": "Посмотреть информацию о модели",
+ "## Download Model": "## Скачать модель",
+ "Model Link": "Ссылка на модель",
+ "Introduce the model link": "Введение ссылки на модель",
+ "Download Model": "Скачать модель",
+ "## Drop files": "## Удаление файлов",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Перетащите файлы .pth и .index в это пространство. Перетащите один, а затем другой.",
+ "TTS Voices": "Голоса TTS",
+ "Select the TTS voice to use for the conversion.": "Выберите голос TTS, который будет использоваться для преобразования.",
+ "Text to Synthesize": "Синтезируемый текст",
+ "Enter the text to synthesize.": "Введите текст для синтеза.",
+ "Or you can upload a .txt file": "Или вы можете загрузить .txt файл",
+ "Enter text to synthesize": "Введите текст для синтеза",
+ "Output Path for TTS Audio": "Выходной тракт для TTS Audio",
+ "Output Path for RVC Audio": "Выходной тракт для RVC Audio",
+ "Enable Applio integration with Discord presence": "Включите интеграцию Applio с присутствием в Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Это активирует возможность отображения текущей активности Applio в Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Включите интеграцию Applio с applio.org/models с помощью flask",
+ "It will activate the possibility of downloading models with a click from the website.": "Он активирует возможность скачивания моделей одним кликом с сайта.",
+ "Theme": "Тема",
+ "Select the theme you want to use. (Requires restarting Applio)": "Выберите тему, которую хотите использовать. (Требуется перезапуск Applio)",
+ "Language": "Язык",
+ "Select the language you want to use. (Requires restarting Applio)": "Выберите язык, который вы хотите использовать. (Требуется перезапуск Applio)",
+ "Plugin Installer": "Установщик плагинов",
+ "Drag your plugin.zip to install it": "Перетащите plugin.zip, чтобы установить его",
+ "Version Checker": "Проверка версий",
+ "Check which version of Applio is the latest to see if you need to update.": "Проверьте, какая версия Applio является последней, чтобы узнать, нужно ли вам обновление.",
+ "Check for updates": "Проверьте наличие обновлений"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/ta-IN.json b/assets/i18n/languages/ta-IN.json
new file mode 100644
index 0000000000000000000000000000000000000000..9da9bdc4320a486a202ddb515a9a680684440f44
--- /dev/null
+++ b/assets/i18n/languages/ta-IN.json
@@ -0,0 +1,204 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "முழுமையான குரல் குளோனிங் கருவி, அநாகரமாக ஒருமிக்க, பகுக்காதது, பயனர் உருவாக்கத்திற்கு உயரியது.\n[தமிழ் மொழிபெயர்ப்பு: Enes](https://discord.com/users/1140031358006202468)",
+ "This section contains some extra utilities that often may be in experimental phases.": "இந்த பிரிவில் சேர்ந்துள்ள கொடுமை கருவிகளில் சார்ந்த பல கூட்டுத்தரங்களைக் கொண்டுள்ளது.",
+ "Output Information": "வெளியீடு தகவல்",
+ "The output information will be displayed here.": "வெளியீடு தகவல் இங்கே காட்டப்படும்.",
+ "Inference": "கருத்து",
+ "Train": "பயிற்சி",
+ "Extra": "கூடுதல்",
+ "Merge Audios": "ஒரேபோனில் ஒன்றாக்குக",
+ "Processing": "செயலாக்கம்",
+ "Audio Analyzer": "ஆடியோ பகுப்பாய்வாளர்",
+ "Model Information": "மாதிரி தகவல்",
+ "Plugins": "பிளகின்கள்",
+ "Download": "பதிவிறக்கம்",
+ "Report a Bug": "பிழை அறிக்கை",
+ "Settings": "அமைப்புகள்",
+ "Preprocess": "முன்பாகவும்",
+ "Model Name": "மாதிரி பெயர்",
+ "Name of the new model.": "புதிய மாதிரி பெயர்.",
+ "Enter model name": "மாதிரி பெயரை உள்ளிடவும்",
+ "Dataset Path": "தரவுத்தொகுதி பாதை",
+ "Path to the dataset folder.": "தரவுத்தொகுதி கோப்புக்கு பாதை.",
+ "Refresh Datasets": "தரவுத்தொகுதிகளை புதுப்பிக்கவும்",
+ "Dataset Creator": "தரவுத்தொகுதி உருவாக்கி",
+ "Dataset Name": "தரவுத்தொகுதி பெயர்",
+ "Name of the new dataset.": "புதிய தரவுத்தொகுதி பெயர்.",
+ "Enter dataset name": "தரவுத்தொகுதி பெயரை உள்ளிடவும்",
+ "Upload Audio Dataset": "ஆடியோ தரவுத்தொகுதியை பதிவேற்றவும்",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ஆடியோ கோப்பு தரவுத்தொகுதிக்கு வெற்றிகரமாக சேர்க்கப்பட்டுள்ளது. தயவுசெய்து முன்னிருப்பை அழுத்தவும்.",
+ "Enter dataset path": "தரவுத்தொகுதி பாதையை உள்ளிடவும்",
+ "Sampling Rate": "மாதிரி விகிதம்",
+ "The sampling rate of the audio files.": "ஆடியோ கோப்புகளின் மாதிரி விகிதம்.",
+ "RVC Version": "RVC பதிப்பு",
+ "The RVC version of the model.": "மாதிரி RVC பதிப்பு.",
+ "Preprocess Dataset": "முன்பாகவும் தரவுத்தொகுதி",
+
+ "Embedder Model": "உள்ளீடு மாதிரி",
+ "Model used for learning speaker embedding.": "பேச்சாளர் உள்ளீட்டை கற்க பயன்படுத்தப்படும் மாதிரி.",
+ "Extract": "எக்ஸ்ட்ராக்ட்",
+ "Hop Length": "ஹாப் நீளம்",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "கருத்துக்கு எவ்வளவு நேரம் எடுத்துக் கொள்கிறது என்றால், அது ஒரு முக்கிய பிச் மாற்றத்திற்கு அமைந்துகொள்கின்றது. சிறிய ஹாப் நீளங்களுக்கு அதிக நேரம் தேவைப்படுகின்றது ஆனால் அவை உயரமான பிச் சரிசெய்தியை உருவாக்க உதவுகின்றன.",
+ "Batch Size": "பேட்ச் அளவு",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "உங்கள் GPU கிடைக்கும் கிடைச்சதை அவசியமாக உள்ளிட பரிந்திருக்கின்றது. 4 என்ற அமைப்பு உயர்த்தப்பட்ட உறுதியுள்ள சொல்லத்தைக் கொண்டுள்ளது ஆனால் அதிக நேரம் பயன்படுகின்றது, 8 அமைப்பு விரைவாக மற்றும் நிலைக்குப் பொருத்தப்படுகிறது.",
+ "Save Every Epoch": "ஒவ்வொரு காலமும் சேமிக்கவும்",
+ "Determine at how many epochs the model will saved at.": "மாதிரி எதிர்காலங்களில் எத்தனை படிப்புகளில் மாதிரியைச் சேமிக்க விரும்புகிறீர்கள்.",
+ "Total Epoch": "மொத்த எபக்",
+ "Specifies the overall quantity of epochs for the model training process.": "மாதிரி பயிற்சி செய்திகளின் மொத்த அளவை குறிப்பிடுகிறது.",
+ "Pretrained": "பூர்வதயாரிக",
+ "Save Only Latest": "கடைசியே சேமிக்கவும்",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "இந்த அமைப்பை இயக்கும் போது G மற்றும் D கோப்புகள் உங்கள் கடைசி பதிப்புகளைச் சேமிக்கும், வாயிலாக சேமிக்கப்படுகின்றன.",
+ "Save Every Weights": "ஒவ்வொரு எடைக்கும் சேமிக்கவும்",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "இந்த அமைப்பு உங்கள் மாதிரி பயிற்சி செய்தியின் முடிவில் மாதிரிகளை சேமிக்க அனுமதிக்கின்றது.",
+ "Custom Pretrained": "கஸ்டம் பூர்வதயாரிக",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "கஸ்டம் பூர்வதயாரிக மாதிரிகளை பயன்படுத்துவது சிறந்த விளக்கங்களை தரலாம், குறிப்பிடுகின்ற குழப்பத்திற்கு ஏற்றதும் பூர்வதயாரிக மாதிரிகளைத் தேர்ந்தெடுக்க உடனே அந்தக் குழப்பத்தை அபூர்வமாக செயல்படுத்தலாம்.",
+ "Upload Pretrained Model": "பூர்வதயாரிக மாதிரி மோடெலை பதிவேற்றவும்",
+ "Refresh Custom Pretraineds": "கஸ்டம் பூர்வதயாரிகளை புதுப்பிக்கவும்",
+ "Pretrained Custom Settings": "கஸ்டம் பூர்வதயாரிக அமைப்புகள்",
+ "The file you dropped is not a valid pretrained file. Please try again.": "நீங்கள் பொருத்தவில்லை என்றால் பூர்வதயாரிக கோப்பு அல்ல. மீண்டும் முயற்சிக்கவும்.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "கீழேயுள்ள பட்டி பட்டியில் பூர்வதயாரிக கோப்புக்கு உருவாக்க முயலுங்கள்.",
+ "Pretrained G Path": "கஸ்டம் பூர்வதயாரிக G பாதை",
+ "Pretrained D Path": "கஸ்டம் பூர்வதயாரிக D பாதை",
+ "GPU Settings": "GPU அமைப்புகள்",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "மேலும் முதிர்ச்சியான GPU அமைப்புகளை அமைக்கின்றது, உடனடியான GPU கருவிக்கு பரிந்திரமான பயனாளர்களுக்கு பரிந்துரிக்கப்படுகிறது.",
+ "GPU Custom Settings": "GPU கஸ்டம் அமைப்புகள்",
+ "GPU Number": "GPU எண்",
+ "0 to ∞ separated by -": "0 இரு ∞ பிரிவாக - வாக்கப்பட்டு",
+ "GPU Information": "GPU தகவல்",
+ "Pitch Guidance": "பிச் வழிநிரப்பல்",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "பிச் வழிநிரப்பல் மூலம், மூல குரலின் ஒலிக்கோட்டைக் கண்டுகொள்வது சாத்தியமாகின்றது, அதன் பிச்சை கூட. இந்த அம்சம் குரல் பாடலுக்கும் மற்றும் உலாவிகளுக்கும் மூல இசை அல்லது பிச் முதிரையைக் காப்பாற்ற எளியதாக இருக்கும்.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "உங்கள் பயிற்சியில் உங்கள் தனிப்பட்ட மாதிரிகளை பயன்படுத்துவது பூர்வதயாரிக மாதிரிகளை பயன்படுத்துவது குரல் பயிற்சி காலத்தை குறைக்கின்றது மற்றும் மொத்த தரவின் உயர்த்துத்தை அதிகரிக்கின்றது.",
+ "Extract Features": "அம்சங்கள் எடு",
+ "Overtraining Detector": "அதிகமாக பயிற்சிப்படுத்தும் அறிவுப்பால்",
+ "Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data.": "பயிற்சிப்படுத்தும் தரவை மிகவும் நன்றாக அறியும் பாதுகாப்பு மூலம் மாதிரி பயிற்சிப்படுத்துதலை தடுக்க, புதிய தரவுக்கு பொதுவாக பொருத்தமாக மாற்ற அனுமதியை இழக்குகிறது.",
+ "Overtraining Detector Settings": "அதிக பயிற்சிப்படுத்தும் அறிவுப்பால் அமைப்புகள்",
+ "Overtraining Threshold": "அதிக பயிற்சிப்படுத்தும் அறிவுப்பால் அதிகம்",
+ "Set the maximum number of epochs you want your model to stop training if no improvement is detected.": "அதிகமாக பயிற்சிப்படுத்தும் தரவு அறியப்படாதால் உங்கள் மாதிரியின் பயிற்சிப்படுத்தும் மொத்த எண்ணிக்கையை அமைக்கவும்.",
+
+ "Start Training": "பயிற்சி ஆரம்பிக்கவும்",
+ "Stop Training & Restart Applio": "பயிற்சி நிறுத்து & புதுப்பிக்க Applio",
+ "Generate Index": "சுருக்கம் உருவாக்கு",
+
+ "Export Model": "ஏற்றுமதி மாதிரி",
+ "The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive.": "'பதிவேற்று' பொத்தான்கள் உள்ளீட்டிற்கு மட்டுமே கூகுள் கோலாப் சேமிப்பகத்திற்கு கடைசியாக கூகுள் டிரைவில் உங்கள் ApplioExported கோப்புக்கு ஏற்றுமதிக்கும்.",
+ "Exported Pth file": "ஏற்றுமதிக்கப்பட்ட Pth கோப்பு",
+ "Exported Index file": "ஏற்றுமதிக்கப்பட்ட சுட்டி கோப்பு",
+ "Select the pth file to be exported": "ஏற்றுமதிக்க வேண்டிய pth கோப்பைத் தேர்ந்தெடுக்கவும்",
+ "Select the index file to be exported": "ஏற்றுமதிக்க வேண்டிய சுட்டி கோப்பைத் தேர்ந்தெடுக்கவும்",
+ "Upload": "பதிவேற்று",
+
+ "Voice Model": "குரல் மாதிரி",
+ "Select the voice model to use for the conversion.": "மாற்றத்திற்கு பயன்படுத்த விரும்பும் குரல் மாதிரியை தேர்ந்தெடுக்கவும்.",
+ "Index File": "சுருக்க கோப்பு",
+ "Select the index file to use for the conversion.": "மாற்றத்திற்கு பயன்படுத்த உள்ள சுருக்க கோப்பை தேர்ந்தெடுக்கவும்.",
+ "Refresh": "புதுப்பிக்கவும்",
+ "Unload Voice": "குரல் அமைதி",
+ "Single": "ஒற்றை",
+ "Upload Audio": "ஒலியை பதிவேற்று",
+ "Select Audio": "ஒலியைத் தேர்ந்தெடு",
+ "Select the audio to convert.": "மாற்றுவதற்கு ஒலியைத் தேர்ந்தெடு.",
+ "Advanced Settings": "மேம்பாடு அமைப்புகள்",
+ "Clear Outputs (Deletes all audios in assets/audios)": "வெற்றிகளை அழித்தல் (assets/audios உள்ள அனைத்து ஒலிகளையும் நீக்கும்)",
+ "Custom Output Path": "கஸ்டம் வெற்றிப் பாதை",
+ "Output Path": "வெற்றி பாதை",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "வெற்றிகள் உள்ளிடப்பட்ட ஒலியைச் சேமிக்கப்படும் பாதை, பொதுவாக assets/audios/output.wav இல்.",
+ "Split Audio": "ஒலியை பிரித்தல்",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "கொலுசுகளாக ஒலியை பிரிக்க, சில நிலைகளில் சிறப்பு விளக்கங்களைப் பெற விரும்புகின்றது.",
+ "Autotune": "ஆட்டோடியூன்",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "உங்கள் முன்னோடிகளுக்கு ஒரு மென்னுரை ஆட்டோடியூனை பயன்படுத்தவும், பாடல் மாற்றங்களுக்கு பரிந்துரிக்கப்படுகின்றது.",
+ "Clean Audio": "சுத்தமான ஒலி",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "உங்கள் ஒலி வெற்றியை சுத்தமாக்க, பேசும் ஒலிகளுக்கு பரிந்துரிக்கப்படுகின்றது.",
+ "Clean Strength": "சுத்த வலிமை",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "நீங்கள் விரும்பும் ஒலிக்கு சுத்தமாக்க விளக்கு, அதை அதிகரிக்கும்போது அது அதிகரிக்கும், ஆனால் ஒலி குறுகியாக இருக்கலாம்.",
+ "Pitch": "பிச்",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ஒலியின் பிச் அமைக்கவும், மதிப்பு உயரானதும் அதிகமாகும்.",
+ "Filter Radius": "குழப்பத்தின் அருகு",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "எண் மூலம் மூன்று அல்லது அதனை விட அதிகமாக இருந்தால், சேகரித்த இசை முடிவுகளில் இயலுமை குறைவாகும் என்று சொல்லப்படுகின்றது.",
+ "Search Feature Ratio": "தேடல் அம்ச விகிதம்",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "இடைவரிசு கோப்பின் மூலம் உள்ள பாதியான ஒருவரிடத்திற்கு உருவாகும் அந்தக் கோப்பு; அதிக மதிப்பு அதிக உருவாகும் என்று அர்த்தம். எனவே, குறோக்கின் குறைந்த மதிப்புகளைத் தேர்வுசெய்வதால் ஒலியில் உள்ள கலப்புகளைத் தவிர்க்க உதவலாம்.",
+ "Volume Envelope": "அளவு என்வெலோப்",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "வெற்றியின் அளவு என்வெலோப் இல் மாறியது அல்லது இணைந்தது. விளக்கு அந்த விகிதம் 1 க்கு அழைத்திருந்தால், வெற்றியின் அளவு என்வெலோப் பயன்படும்.",
+ "Protect Voiceless Consonants": "குரலின் இல்லாத ஸ்வரக்களைக் காப்பாற்றவும்",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "எலக்ட்ரோ-ஒலி கொழுகு மற்றும் பிற கலப்புகளை தடுக்குவதற்கு விரிவான ஸ்வரக்களுக்கு மற்றும் சுவாசத் தானங்களுக்கு பாதுகாக்க. இந்த அளவுக்கு அதிகமாக 0.5 க்கு அழைத்துக் கொள்வது பொருத்தமான பாதுகாப்பை வழங்குகின்றது. ஆனால், இந்த மதிப்பை குறைந்ததாக்கின்றார் என்றால், பாதுகாப்புக்குரிய நிலை குறைந்துவிடப்படலாம் மற்றும் அதுவே இந்தக் குறோக்குனை பரிந்துரிக்கும் என்று எச்சரிக்கை தரகின்றது.",
+ "Pitch extraction algorithm": "பிச் எக்ஸ்டிரக்ஷன் அளவுத் தொகுப்பு",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ஒலி மாற்றத்திற்கு பயன்படுத்த வேண்டிய பிச் எக்ஸ்டிரக்ஷன் அளவுத் தொகுப்பு. இயல்பான அளவுத் தொகுப்பு rmvpe ஆகும், இது அதிகமாக பரிந்துரிக்கப்படுகின்றது.",
+
+ "Convert": "மாற்று",
+ "Export Audio": "ஒலியை ஏற்றுமதி செய்",
+
+ "Batch": "பேட்ச்",
+ "Input Folder": "உள்ளிடும் கோப்பு",
+ "Select the folder containing the audios to convert.": "மாற்ற ஒலிகளைக் கொண்ட கோப்புகளைக் கொண்ட கோப்புக்கு தேர்ந்தெடு.",
+ "Enter input path": "உள்ளிடும் பாதையை உள்ளிடுக",
+ "Output Folder": "வெற்றி கோப்பு",
+ "Select the folder where the output audios will be saved.": "வெற்றிகளைச் சேமிக்கப்படும் கோப்புக்கு தேர்ந்தெடு.",
+ "Enter output path": "வெற்றியின் பாதையை உள்ளிடுக",
+
+ "Get information about the audio": "ஒலியை பற்றிய தகவல் பெறுக",
+
+ "## Voice Blender": "## குரல் பிளெண்டர்",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "இரண்டு குரல் மாதிரிகளைத் தேர்வு செய்து, விரும்பிய குரல் சதவீதம் அமைக்கவும், அவைகளை முழுமையாக ஒரு புதிய குரலாக பிளெண்டுகின்றன.",
+ "Voice Blender": "குரல் பிளெண்டர்",
+ "Drag and drop your model here": "உங்கள் மாதிரி இங்கே எழுதவும்",
+ "You can also use a custom path.": "நீங்கள் கஸ்டம் பாதையையும் பயன்படுத்தலாம்.",
+ "Blend Ratio": "குரல் சதவீதம்",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "ஒரு பக்கத்திற்கு அல்லது மற்றும் மற்றும் அதிகமாக செய்யும் உள்ளீடு இரட்டிப் பார்த்துக் கொள்ளுதல் மாதிரியாகின்றது.",
+ "Fusion": "ஐக்கியம்",
+ "Path to Model": "மாதிரிக்கு பாதை",
+ "Enter path to model": "மாதிரிக்கு பாதையை உள்ளிடுக",
+ "Model information to be placed": "இருந்து விடப்பட வேண்டிய மாதிரி தகவல்",
+ "Introduce the model information": "மாதிரி தகவல் அறிமுகம்",
+ "The information to be placed in the model (You can leave it blank or put anything).": "மாதிரிக்கு வைக்கப்பட வேண்டிய தகவல் (நீங்கள் அந்தச் செயலை விட அந்தச் செய்யாமல் அனைத்ததையும் வைக்கலாம்).",
+ "View model information": "மாதிரி தகவலைக் காண்க",
+ "Introduce the model pth path": "மாதிரி pth பாதையை உள்ளிடுக",
+ "View": "காண்க",
+ "Model extraction": "மாதிரி எடுத்தல்",
+ "Model conversion": "மாதிரி மாற்றம்",
+ "Pth file": "Pth கோப்பு",
+ "Output of the pth file": "Pth கோப்பின் வெளியேற்றம்",
+ "# How to Report an Issue on GitHub": "# GitHub-ல் ஒரு பிரச்சினையை புகாரளிக்குவது எப்படி",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. நீங்கள் அந்தப் பிரச்சினையை பரிசோதிக்கும் கீழே 'திரையை பதிகம் செய்யும்' பொத்தானை கிளிக் செய்யவும்.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. நீங்கள் அந்தப் பிரச்சினையை பரிசோதித்துக் கொண்டிருக்கின்றீர்கள், அந்தச் செய்யப்படும் பொத்தானை கிளிக் செய்யவும் (இது நீங்கள் சொல்லப்படும் பொத்தான், ஆனால் நீங்கள் எந்தவேணையும் செய்யக்கூடிய நிலையின் போது பொத்தானின் பெயர் மாறுகின்றது).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub Issues](https://github.com/IAHispano/Applio/issues) க்கு செல்லவும் மற்றும் 'புதிய பிரச்சினை' பொத்தானை கிளிக் செய்யவும்.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. வழுதுணர்ந்து, தேவையான விவரங்களைக் கொண்டு விரிவாக பிரச்சினை பதிவேடு செய்து, முந்தைய படித்த கோப்பை பதிவேடுக்கு பயன்படுத்தலாம்.",
+ "Record Screen": "திரையை பதிகம் செய்க",
+ "Record": "பதிகம் செய்க",
+ "Stop Recording": "பதிகம் நிறுத்துக",
+ "Introduce the model .pth path": "மாதிரி .pth பாதையை உள்ளிடுக",
+ "See Model Information": "மாதிரி தகவலைக் காண்க",
+ "## Download Model": "## மாதிரி பதிவிறக்கம்",
+ "Model Link": "மாதிரி இணைப்பு",
+ "Introduce the model link": "மாதிரி இணைப்பை உள்ளிடுக",
+ "Download Model": "மாதிரி பதிவிறக்கம்",
+ "## Drop files": "## கோப்புகளை விழுக",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "உங்கள் .pth கோப்பு மற்றும் .index கோப்பை இந்த இடத்திற்கு எழுதுங்கள். ஒருவருக்கு பிறகு ஒருவருக்கு எழுதுங்கள்.",
+ "## Search Model": "## மாதிரி தேடு",
+ "Search": "தேடு",
+ "Introduce the model name to search.": "தேடுவதற்கு மாதிரிப் பெயரை அறிமுகப்படுத்து.",
+ "We couldn't find models by that name.": "அந்த பெயரில் மாதிரிகளைக் கண்டுபிடிக்க முடியவில்லை.",
+
+ "TTS Voices": "TTS குரல்கள்",
+ "Select the TTS voice to use for the conversion.": "மாற்றத்திற்கு பயன்படுத்த உள்ள TTS குரலை தேர்ந்தெடுக்கவும்.",
+ "Text to Synthesize": "சிந்தனை செய்ய உள்ள உரை",
+ "Enter the text to synthesize.": "சிந்தனை செய்ய உள்ள உரையை உள்ளிடுக.",
+ "Or you can upload a .txt file": "அல்லது .txt கோப்பை பதிவேற்றலாம்",
+ "Enter text to synthesize": "சிந்தனை செய்ய உள்ள உரையை உள்ளிடுக",
+ "Output Path for TTS Audio": "TTS குரலுக்கான வெளியேற்ற பாதை",
+ "Output Path for RVC Audio": "RVC குரலுக்கான வெளியேற்ற பாதை",
+ "Enable Applio integration with Discord presence": "Discord உள்ளிட்டது உள்ளிடத்துடன் Applio ஒருவருக்கு இயங்குதல் இயல்புநிலை செய்தியை இயக்குங்கள்",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "இது Applio செயல்திறனை Discord-ல் காண்பிக்க முடியும்.",
+ "Enable Applio integration with applio.org/models using flask": "flask ஐப் பயன்படுத்தி applio.org/models உடன் Applio ஒருவருக்கு இயங்குதல் இயல்புநிலை செய்தியை இயக்குங்கள்",
+ "It will activate the possibility of downloading models with a click from the website.": "இது இணையத்திலிருந்து ஒரு கிளிக்குட்டுடன் மாதிரிகளை பதிவிறக்க முடியும்.",
+ "Enable fake GPU": "கப்பூ ஜி.பி.யூ ஐ இயக்கு",
+ "Training is currently unsupported due to the absence of a GPU. To activate the training tab, navigate to the settings tab and enable the 'Fake GPU' option.": "ஒரு ஜி.பி.யூ இல்லாமல் தற்போது பயிற்சிப்படுத்தல் ஆதரிக்கப்படவில்லை. பயிற்சிக்கு பட்டி செயல்முறையை செயலாக்க, அமைப்புகள் பட்டியலில் செல்ல, 'பெயர் ஜி.பி.யூ' விருப்பத்தை இயக்கவும்.",
+ "Activates the train tab. However, please note that this device lacks GPU capabilities, hence training is not supported. This option is only for testing purposes. (This option will restart Applio)": "பயிற்சிக்கு பட்டி செயலாக்கிறது. எனவே, இந்த சாதனம் ஜி.பி.யூ திறன் இல்லையாம், அதனால் பயிற்சி ஆதரிக்கப்படவில்லை. இந்த விருப்பம் மட்டுமே சோதனை காரணங்களுக்காக உள்ளது. (இந்த விருப்பம் Applio ஐ மீண்டும் திரும்பியிருப்பதற்காக)",
+ "Theme": "தீமா",
+ "Select the theme you want to use. (Requires restarting Applio)": "நீங்கள் பயன்படுத்த விரும்பும் தீமையை தேர்ந்தெடுக்கவும். (Applio-ஐ மீளவே போகும்)",
+ "Language": "மொழி",
+ "Select the language you want to use. (Requires restarting Applio)": "நீங்கள் பயன்படுத்த விரும்பும் மொழியை தேர்ந்தெடுக்கவும். (Applio-ஐ மீளவே போகும்)",
+ "Plugin Installer": "பிளகின் நிறுவி",
+ "Drag your plugin.zip to install it": "உங்கள் plugin.zip கோப்பை இதுக்கு இழுக்கவும் அதை நிறுவுக",
+ "Version Checker": "பதிப்பு சரிபார்க்கல்",
+ "Check which version of Applio is the latest to see if you need to update.": "நீங்கள் புதியதாகப் புதுப்பிக்க வேண்டும் என்பதை பார்க்க, Applio இன் எந்த பதிப்பு சரிபார்க்கவும்.",
+ "Check for updates": "புதுப்பிக்கவும்"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/ta_TA.json b/assets/i18n/languages/ta_TA.json
new file mode 100644
index 0000000000000000000000000000000000000000..6dd0b5908c041b11a57dde93c2f4a54dc0e2697d
--- /dev/null
+++ b/assets/i18n/languages/ta_TA.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "இறுதி குரல் குளோனிங் கருவி, நிகரற்ற சக்தி, மாடுலாரிட்டி மற்றும் பயனர் நட்பு அனுபவத்திற்காக உன்னிப்பாக உகந்ததாக உள்ளது.",
+ "This section contains some extra utilities that often may be in experimental phases.": "இந்த பிரிவில் சில கூடுதல் பயன்பாடுகள் உள்ளன, அவை பெரும்பாலும் சோதனை கட்டங்களில் இருக்கலாம்.",
+ "Output Information": "வெளியீட்டு தகவல்",
+ "The output information will be displayed here.": "வெளியீட்டு தகவல் இங்கே காட்டப்படும்.",
+ "Inference": "அனுமானம்",
+ "Train": "தொடர்வண்டி",
+ "Extra": "கூடுதல்",
+ "Merge Audios": "ஆடியோக்களை ஒன்றிணைக்கவும்",
+ "Processing": "செயலாக்க",
+ "Audio Analyzer": "ஆடியோ அனலைசர்",
+ "Model Information": "மாதிரி தகவல்",
+ "Plugins": "செருகுநிரல்கள்",
+ "Download": "பதிவிறக்க",
+ "Report a Bug": "பிழையைப் புகாரளி",
+ "Settings": "அமைப்புகள்",
+ "Preprocess": "முன் செயல்முறை",
+ "Model Name": "மாடல் பெயர்",
+ "Name of the new model.": "புதிய மாடலின் பெயர்.",
+ "Enter model name": "மாடல் பெயரை உள்ளிடவும்",
+ "Dataset Path": "தரவுத்தொகுப்பு பாதை",
+ "Path to the dataset folder.": "தரவுத்தொகுப்பு கோப்புறைக்கான பாதை.",
+ "Refresh Datasets": "தரவுத்தொகுப்புகளைப் புதுப்பிக்கவும்",
+ "Dataset Creator": "தரவுத்தொகுப்பை உருவாக்கியவர்",
+ "Dataset Name": "தரவுத்தொகுப்பு பெயர்",
+ "Name of the new dataset.": "புதிய தரவுத்தொகுப்பின் பெயர்.",
+ "Enter dataset name": "தரவுத்தொகுப்பு பெயரை உள்ளிடவும்",
+ "Upload Audio Dataset": "ஆடியோ தரவுத்தொகுப்பைப் பதிவேற்றவும்",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ஆடியோ கோப்பு வெற்றிகரமாக தரவுத்தொகுப்பில் சேர்க்கப்பட்டது. செயல்முறைக்கு முந்தைய பட்டனை கிளிக் செய்யவும்.",
+ "Enter dataset path": "தரவுத்தொகுப்பு பாதையை உள்ளிடவும்",
+ "Sampling Rate": "மாதிரி மதிப்பீடு",
+ "The sampling rate of the audio files.": "ஆடியோ கோப்புகளின் மாதிரி விகிதம்.",
+ "RVC Version": "RVC பதிப்பு",
+ "The RVC version of the model.": "மாடலின் RVC பதிப்பு.",
+ "Preprocess Dataset": "செயல்முறைக்கு முந்தைய தரவுத்தொகுப்பு",
+ "Extract": "பிரித்தெடுத்தல்",
+ "Hop Length": "ஹாப் நீளம்",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "கணினி ஒரு குறிப்பிடத்தக்க சுருதி மாற்றத்திற்கு மாறுவதற்கு எடுக்கும் காலத்தைக் குறிக்கிறது. சிறிய ஹாப் நீளங்களுக்கு அனுமானத்திற்கு அதிக நேரம் தேவைப்படுகிறது, ஆனால் அதிக சுருதி துல்லியத்தை அளிக்க முனைகிறது.",
+ "Batch Size": "தொகுதி அளவு",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "உங்கள் GPU இன் கிடைக்கக்கூடிய VRAM உடன் அதை சீரமைப்பது நல்லது. 4 இன் அமைப்பு மேம்பட்ட துல்லியம் ஆனால் மெதுவான செயலாக்கத்தை வழங்குகிறது, அதே நேரத்தில் 8 வேகமான மற்றும் நிலையான முடிவுகளை வழங்குகிறது.",
+ "Save Every Epoch": "ஒவ்வொரு சகாப்தத்தையும் காப்பாற்றுங்கள்",
+ "Determine at how many epochs the model will saved at.": "மாதிரி எத்தனை சகாப்தங்களில் சேமிக்கப்படும் என்பதை தீர்மானிக்கவும்.",
+ "Total Epoch": "மொத்த சகாப்தம்",
+ "Specifies the overall quantity of epochs for the model training process.": "மாதிரி பயிற்சி செயல்முறைக்கான சகாப்தங்களின் ஒட்டுமொத்த அளவைக் குறிப்பிடுகிறது.",
+ "Pretrained": "முன் பயிற்சி",
+ "Save Only Latest": "சமீபத்தியதை மட்டும் சேமிக்கவும்",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "இந்த அமைப்பை இயக்குவது G மற்றும் D கோப்புகளை அவற்றின் மிகச் சமீபத்திய பதிப்புகளை மட்டுமே சேமிக்கும், சேமிப்பக இடத்தை திறம்பட சேமிக்கும்.",
+ "Save Every Weights": "ஒவ்வொரு எடைகளையும் சேமிக்கவும்",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "இந்த அமைப்பு ஒவ்வொரு சகாப்தத்தின் முடிவிலும் மாதிரியின் எடைகளை சேமிக்க உதவுகிறது.",
+ "Custom Pretrained": "தனிப்பயன் முன்பயிற்சி",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "தனிப்பயன் முன்பயிற்சி மாதிரிகளைப் பயன்படுத்துவது சிறந்த முடிவுகளுக்கு வழிவகுக்கும், ஏனெனில் குறிப்பிட்ட பயன்பாட்டு வழக்குக்கு ஏற்ப மிகவும் பொருத்தமான முன்பயிற்சி மாதிரிகளைத் தேர்ந்தெடுப்பது செயல்திறனை கணிசமாக மேம்படுத்தும்.",
+ "Upload Pretrained Model": "முன்பயிற்சி பெற்ற மாதிரியைப் பதிவேற்றவும்",
+ "Refresh Custom Pretraineds": "தனிப்பயன் ப்ரீட்ரெய்ன்ட்களைப் புதுப்பிக்கவும்",
+ "Pretrained Custom Settings": "முன் பயிற்சி பெற்ற தனிப்பயன் அமைப்புகள்",
+ "The file you dropped is not a valid pretrained file. Please try again.": "நீங்கள் கைவிட்ட கோப்பு ஒரு செல்லத்தக்க முன்பயிற்சி பெற்ற கோப்பு அல்ல. மீண்டும் முயற்சிக்கவும்.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "கீழ்தோன்றும் மெனுவில் முன்பயிற்சி பெற்ற கோப்பைக் காண புதுப்பிப்பு பொத்தானைக் கிளிக் செய்யவும்.",
+ "Pretrained G Path": "தனிப்பயன் முன் பயிற்சி ஜி",
+ "Pretrained D Path": "தனிப்பயன் முன்பயிற்சி D",
+ "GPU Settings": "GPU அமைப்புகள்",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "மேம்பட்ட GPU அமைப்புகளை அமைக்கிறது, சிறந்த GPU கட்டமைப்பு கொண்ட பயனர்களுக்கு பரிந்துரைக்கப்படுகிறது.",
+ "GPU Custom Settings": "GPU தனிப்பயன் அமைப்புகள்",
+ "GPU Number": "GPU எண்",
+ "0 to ∞ separated by -": "0 முதல் ∞ வரை பிரிக்கப்பட்டுள்ளது -",
+ "GPU Information": "தகவல்",
+ "Pitch Guidance": "சுருதி வழிகாட்டுதல்",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "சுருதி வழிகாட்டுதலைப் பயன்படுத்துவதன் மூலம், அதன் சுருதி உட்பட அசல் குரலின் தொனியை பிரதிபலிப்பது சாத்தியமாகும். அசல் மெல்லிசை அல்லது சுருதி வடிவத்தைப் பாதுகாப்பது அவசியம் என்று பாடுவதற்கும் பிற காட்சிகளுக்கும் இந்த அம்சம் குறிப்பாக மதிப்புமிக்கது.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "உங்கள் சொந்த பயிற்சியின் போது முன்கூட்டியே பயிற்சி பெற்ற மாதிரிகளைப் பயன்படுத்தவும். இந்த அணுகுமுறை பயிற்சி காலத்தை குறைக்கிறது மற்றும் ஒட்டுமொத்த தரத்தை மேம்படுத்துகிறது.",
+ "Extract Features": "பிரித்தெடுக்கும் அம்சங்கள்",
+ "Start Training": "பயிற்சியைத் தொடங்குங்கள்",
+ "Generate Index": "குறியீட்டை உருவாக்கவும்",
+ "Voice Model": "குரல் மாதிரி",
+ "Select the voice model to use for the conversion.": "மாற்றத்திற்கு பயன்படுத்த குரல் மாதிரியைத் தேர்ந்தெடுக்கவும்.",
+ "Index File": "அட்டவணை கோப்பு",
+ "Select the index file to use for the conversion.": "மாற்றத்திற்கு பயன்படுத்த குறியீட்டு கோப்பைத் தேர்ந்தெடுக்கவும்.",
+ "Refresh": "புதுப்பி",
+ "Unload Voice": "குரலை இறக்கவும்",
+ "Single": "ஒற்றை",
+ "Upload Audio": "ஆடியோவை பதிவேற்றவும்",
+ "Select Audio": "ஆடியோவை தேர்ந்தெடு",
+ "Select the audio to convert.": "மாற்ற ஆடியோவைத் தேர்ந்தெடுக்கவும்.",
+ "Advanced Settings": "மேம்பட்ட அமைப்புகள்",
+ "Clear Outputs (Deletes all audios in assets/audios)": "வெளியீடுகளை அழிக்கவும் (சொத்துக்கள் / ஆடியோக்களில் உள்ள அனைத்து ஆடியோக்களையும் நீக்குகிறது)",
+ "Custom Output Path": "தனிப்பயன் வெளியீட்டு பாதை",
+ "Output Path": "வெளியீட்டுப் பாதை",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "வெளியீட்டு ஆடியோ சேமிக்கப்படும் பாதை, இயல்பாக சொத்துக்கள் / ஆடியோக்கள் / output.wav",
+ "Split Audio": "பிளவு ஆடியோ",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "சில சந்தர்ப்பங்களில் சிறந்த முடிவுகளைப் பெற அனுமானத்திற்காக ஆடியோவை துண்டுகளாகப் பிரிக்கவும்.",
+ "Autotune": "ஆட்டோடியூன்",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "உங்கள் அனுமானங்களுக்கு ஒரு மென்மையான ஆட்டோட்யூனைப் பயன்படுத்துங்கள், பாடல் மாற்றங்களுக்கு பரிந்துரைக்கப்படுகிறது.",
+ "Clean Audio": "சுத்தமான ஆடியோ",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "ஆடியோக்களைப் பேசுவதற்குப் பரிந்துரைக்கப்படும் இரைச்சல் கண்டறிதல் அல்காரிதம்களைப் பயன்படுத்தி உங்கள் ஆடியோ அவுட்புட்டை சுத்தம் செய்யவும்.",
+ "Clean Strength": "சுத்தமான வலிமை",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "நீங்கள் விரும்பும் ஆடியோவுக்கு சுத்தம் செய்யும் அளவை அமைக்கவும், நீங்கள் அதை எவ்வளவு அதிகரிக்கிறீர்களோ, அவ்வளவு அதிகமாக அது சுத்தம் செய்யப்படும், ஆனால் ஆடியோ மிகவும் சுருக்கப்பட்டிருக்கும்.",
+ "Pitch": "எறி",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ஆடியோவின் சுருதியை அமைக்கவும், அதிக மதிப்பு, அதிக சுருதி.",
+ "Filter Radius": "வடிகட்டி ஆரம்",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "எண்ணிக்கை மூன்றை விட அதிகமாகவோ அல்லது சமமாகவோ இருந்தால், சேகரிக்கப்பட்ட தொனி முடிவுகளில் சராசரி வடிகட்டலைப் பயன்படுத்துவது சுவாசத்தைக் குறைக்கும் ஆற்றலைக் கொண்டுள்ளது.",
+ "Search Feature Ratio": "தேடல் அம்ச விகிதம்",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "குறியீட்டு கோப்பு செலுத்தும் செல்வாக்கு; அதிக மதிப்பு அதிக செல்வாக்குக்கு ஒத்திருக்கிறது. இருப்பினும், குறைந்த மதிப்புகளைத் தேர்ந்தெடுப்பது ஆடியோவில் இருக்கும் கலைப்பொருட்களைத் தணிக்க உதவும்.",
+ "Volume Envelope": "தொகுதி உறை",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "வெளியீட்டின் தொகுதி உறையுடன் மாற்றவும் அல்லது கலக்கவும். விகிதம் 1 க்கு நெருக்கமாக இருந்தால், வெளியீடு உறை அதிகமாக பயன்படுத்தப்படுகிறது.",
+ "Protect Voiceless Consonants": "குரலற்ற மெய்யெழுத்துக்களைப் பாதுகாக்கவும்",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "மின்-ஒலி கிழித்தல் மற்றும் பிற கலைப்பொருட்களைத் தடுக்க தனித்துவமான மெய்யெழுத்துக்கள் மற்றும் சுவாச ஒலிகளைப் பாதுகாக்கவும். அளவுருவை அதன் அதிகபட்ச மதிப்பான 0.5 க்கு இழுப்பது விரிவான பாதுகாப்பை வழங்குகிறது. இருப்பினும், இந்த மதிப்பைக் குறைப்பது பாதுகாப்பின் அளவைக் குறைக்கலாம், அதே நேரத்தில் குறியீட்டு விளைவைத் தணிக்கலாம்.",
+ "Pitch extraction algorithm": "சுருதி பிரித்தெடுத்தல் அல்காரிதம்",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ஆடியோ மாற்றத்திற்கு பயன்படுத்த சுருதி பிரித்தெடுத்தல் வழிமுறை. இயல்புநிலை அல்காரிதம் rmvpe ஆகும், இது பெரும்பாலான சந்தர்ப்பங்களில் பரிந்துரைக்கப்படுகிறது.",
+ "Convert": "உருமாற்று",
+ "Export Audio": "ஆடியோவை ஏற்றுமதி செய்யவும்",
+ "Batch": "தொகுதி",
+ "Input Folder": "உள்ளீட்டு கோப்புறை",
+ "Select the folder containing the audios to convert.": "மாற்ற ஆடியோக்களைக் கொண்ட கோப்புறையைத் தேர்ந்தெடுக்கவும்.",
+ "Enter input path": "உள்ளீட்டு பாதையை உள்ளிடவும்",
+ "Output Folder": "வெளியீட்டு கோப்புறை",
+ "Select the folder where the output audios will be saved.": "வெளியீடு ஆடியோக்கள் சேமிக்கப்படும் கோப்புறையைத் தேர்ந்தெடுக்கவும்.",
+ "Enter output path": "வெளியீட்டு பாதையை உள்ளிடவும்",
+ "Get information about the audio": "ஆடியோ பற்றிய தகவலைப் பெறுங்கள்",
+ "Information about the audio file": "ஆடியோ கோப்பு பற்றிய தகவல்",
+ "Waiting for information...": "தகவலுக்காக காத்திருக்கிறேன்...",
+ "## Voice Blender": "## வாய்ஸ் பிளெண்டர்",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "இரண்டு குரல் மாதிரிகளைத் தேர்ந்தெடுத்து, நீங்கள் விரும்பிய கலவை சதவீதத்தை அமைத்து, அவற்றை முற்றிலும் புதிய குரலில் கலக்கவும்.",
+ "Voice Blender": "குரல் பிளெண்டர்",
+ "Drag and drop your model here": "உங்கள் மாதிரியை இங்கே இழுத்து விடுங்கள்",
+ "You can also use a custom path.": "நீங்கள் தனிப்பயன் பாதையையும் பயன்படுத்தலாம்.",
+ "Blend Ratio": "கலப்பு விகிதம்",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "நிலையை ஒரு பக்கம் அல்லது மறுபுறம் நோக்கி சரிசெய்வது மாதிரியை முதல் அல்லது இரண்டாவதைப் போலவே மாற்றும்.",
+ "Fusion": "இணைவு",
+ "Path to Model": "மாதிரிக்கான பாதை",
+ "Enter path to model": "மாதிரிக்கான பாதையை உள்ளிடவும்",
+ "Model information to be placed": "வைக்கப்பட வேண்டிய மாதிரி தகவல்",
+ "Inroduce the model information": "மாதிரி தகவலை அறிமுகப்படுத்தவும்",
+ "The information to be placed in the model (You can leave it blank or put anything).": "மாதிரியில் வைக்கப்பட வேண்டிய தகவல் (நீங்கள் அதை காலியாக விடலாம் அல்லது எதையும் வைக்கலாம்).",
+ "View model information": "மாதிரி தகவலைக் காண்க",
+ "Introduce the model pth path": "மாதிரி pth பாதையை அறிமுகப்படுத்தவும்",
+ "View": "பார்வை",
+ "Model extraction": "மாதிரி பிரித்தெடுத்தல்",
+ "Model conversion": "மாதிரி மாற்றம்",
+ "Pth file": "Pth கோப்பு",
+ "Output of the pth file": "pth கோப்பின் வெளியீடு",
+ "# How to Report an Issue on GitHub": "# GitHub இல் ஒரு சிக்கலை எவ்வாறு புகாரளிப்பது",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. நீங்கள் அனுபவிக்கும் சிக்கலைப் பதிவு செய்யத் தொடங்க கீழே உள்ள 'ரெக்கார்ட் ஸ்கிரீன்' பொத்தானைக் கிளிக் செய்க.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. நீங்கள் சிக்கலைப் பதிவு செய்து முடித்ததும், 'பதிவு செய்வதை நிறுத்து' பொத்தானைக் கிளிக் செய்க (அதே பொத்தான், ஆனால் நீங்கள் தீவிரமாக பதிவு செய்கிறீர்களா இல்லையா என்பதைப் பொறுத்து லேபிள் மாறுகிறது).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub Issues](https://github.com/IAHispano/Applio/issues) என்பதற்குச் சென்று 'புதிய சிக்கல்' பொத்தானைக் கிளிக் செய்யவும்.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. வழங்கப்பட்ட சிக்கல் வார்ப்புருவை முடிக்கவும், தேவைக்கேற்ப விவரங்களைச் சேர்ப்பதை உறுதிசெய்து, முந்தைய கட்டத்திலிருந்து பதிவுசெய்யப்பட்ட கோப்பை பதிவேற்ற சொத்துக்கள் பிரிவைப் பயன்படுத்தவும்.",
+ "Record Screen": "பதிவு திரை",
+ "Record": "பதிவேடு",
+ "Stop Recording": "பதிவு செய்வதை நிறுத்து",
+ "Introduce the model .pth path": "மாதிரியை அறிமுகப்படுத்துங்கள் .pth பாதையை அறிமுகப்படுத்துங்கள்",
+ "See Model Information": "மாதிரி தகவலைப் பார்க்கவும்",
+ "## Download Model": "## பதிவிறக்க மாதிரி",
+ "Model Link": "மாதிரி இணைப்பு",
+ "Introduce the model link": "மாதிரி இணைப்பை அறிமுகப்படுத்தவும்",
+ "Download Model": "மாடலைப் பதிவிறக்கவும்",
+ "## Drop files": "## கோப்புகளை கைவிடுங்கள்",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "உங்கள் .pth கோப்பு மற்றும் .index கோப்பை இந்த இடத்திற்கு இழுக்கவும். ஒன்றை இழுத்து மற்றொன்றை இழுக்கவும்.",
+ "TTS Voices": "TTS குரல்கள்",
+ "Select the TTS voice to use for the conversion.": "மாற்றத்திற்கு பயன்படுத்த TTS குரலைத் தேர்ந்தெடுக்கவும்.",
+ "Text to Synthesize": "தொகுக்க உரை",
+ "Enter the text to synthesize.": "ஒருங்கிணைக்க உரையை உள்ளிடவும்.",
+ "Or you can upload a .txt file": "அல்லது .txt கோப்பை பதிவேற்றலாம்",
+ "Enter text to synthesize": "ஒருங்கிணைக்க உரையை உள்ளிடவும்",
+ "Output Path for TTS Audio": "TTS ஆடியோவுக்கான வெளியீட்டு பாதை",
+ "Output Path for RVC Audio": "RVC ஆடியோவுக்கான வெளியீட்டு பாதை",
+ "Enable Applio integration with Discord presence": "Discord இருப்புடன் Applio ஒருங்கிணைப்பை இயக்கவும்",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "டிஸ்கார்டில் தற்போதைய Applio செயல்பாட்டைக் காண்பிப்பதற்கான வாய்ப்பை இது செயல்படுத்தும்.",
+ "Enable Applio integration with applio.org/models using flask": "குடுவையைப் பயன்படுத்தி applio.org/models உடன் அப்லியோ ஒருங்கிணைப்பை இயக்கவும்",
+ "It will activate the possibility of downloading models with a click from the website.": "இணையதளத்தில் இருந்து ஒரு கிளிக்கில் மாடல்களைப் பதிவிறக்கும் வாய்ப்பை இது செயல்படுத்தும்.",
+ "Theme": "கருப்பொருள்",
+ "Select the theme you want to use. (Requires restarting Applio)": "நீங்கள் பயன்படுத்த விரும்பும் கருப்பொருளை தேர்ந்தெடுக்கவும். (அப்ளியோவை மறுதொடக்கம் செய்ய வேண்டும்)",
+ "Language": "மொழி",
+ "Select the language you want to use. (Requires restarting Applio)": "நீங்கள் பயன்படுத்த விரும்பும் மொழியைத் தேர்ந்தெடுக்கவும். (அப்ளியோவை மறுதொடக்கம் செய்ய வேண்டும்)",
+ "Plugin Installer": "செருகுநிரல் நிறுவி",
+ "Drag your plugin.zip to install it": "அதை நிறுவ உங்கள் plugin.zip இழுக்கவும்",
+ "Version Checker": "பதிப்பு சரிபார்ப்பு",
+ "Check which version of Applio is the latest to see if you need to update.": "நீங்கள் புதுப்பிக்க வேண்டுமா என்பதைப் பார்க்க அப்லியோவின் எந்த பதிப்பு சமீபத்தியது என்பதைச் சரிபார்க்கவும்.",
+ "Check for updates": "புதுப்பிப்புகளைச் சரிபார்க்கவும்"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/te_TE.json b/assets/i18n/languages/te_TE.json
new file mode 100644
index 0000000000000000000000000000000000000000..e82fb12309d966d68fefbf561cc788e705749a80
--- /dev/null
+++ b/assets/i18n/languages/te_TE.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "అల్టిమేట్ వాయిస్ క్లోనింగ్ టూల్, సాటిలేని శక్తి, మాడ్యులారిటీ మరియు వినియోగదారు-స్నేహపూర్వక అనుభవం కోసం జాగ్రత్తగా ఆప్టిమైజ్ చేయబడింది.",
+ "This section contains some extra utilities that often may be in experimental phases.": "ఈ విభాగంలో కొన్ని అదనపు ఉపయోగాలు ఉన్నాయి, అవి తరచుగా ప్రయోగాత్మక దశలలో ఉండవచ్చు.",
+ "Output Information": "అవుట్ పుట్ సమాచారం",
+ "The output information will be displayed here.": "అవుట్ పుట్ సమాచారం ఇక్కడ ప్రదర్శించబడుతుంది.",
+ "Inference": "అనిమితి",
+ "Train": "రైలు",
+ "Extra": "అదనం",
+ "Merge Audios": "ఆడియోలను విలీనం చేయండి",
+ "Processing": "ప్రాసెసింగ్",
+ "Audio Analyzer": "Audio Analyzer",
+ "Model Information": "నమూనా సమాచారం",
+ "Plugins": "Plugins",
+ "Download": "డౌన్ లోడ్ చేసుకోండి",
+ "Report a Bug": "బగ్ ని రిపోర్ట్ చేయండి",
+ "Settings": "సెట్టింగ్ లు",
+ "Preprocess": "ప్రీప్రాసెస్",
+ "Model Name": "మోడల్ పేరు",
+ "Name of the new model.": "కొత్త మోడల్ పేరు..",
+ "Enter model name": "మోడల్ పేరు నమోదు చేయండి",
+ "Dataset Path": "Dataset Path",
+ "Path to the dataset folder.": "డేటాసెట్ ఫోల్డర్ కు మార్గం.",
+ "Refresh Datasets": "రిఫ్రెష్ డేటాసెట్ లు",
+ "Dataset Creator": "డేటాసెట్ సృష్టికర్త",
+ "Dataset Name": "డేటాసెట్ పేరు",
+ "Name of the new dataset.": "కొత్త డేటాసెట్ పేరు.",
+ "Enter dataset name": "డేటాసెట్ పేరును నమోదు చేయండి",
+ "Upload Audio Dataset": "ఆడియో డేటాసెట్ అప్ లోడ్ చేయండి",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "ఆడియో ఫైల్ డేటాసెట్ కు విజయవంతంగా జోడించబడింది. దయచేసి ప్రీప్రాసెస్ బటన్ మీద క్లిక్ చేయండి.",
+ "Enter dataset path": "డేటాసెట్ మార్గాన్ని నమోదు చేయండి",
+ "Sampling Rate": "నమూనా రేటు",
+ "The sampling rate of the audio files.": "ఆడియో ఫైల్స్ యొక్క నమూనా రేటు.",
+ "RVC Version": "RVC Version",
+ "The RVC version of the model.": "మోడల్ యొక్క ఆర్ విసి వెర్షన్.",
+ "Preprocess Dataset": "ప్రీప్రాసెస్ Dataset",
+ "Extract": "ఉద్ధరించు",
+ "Hop Length": "హాప్ పొడవు",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "గణనీయమైన పిచ్ మార్పుకు సిస్టమ్ పరివర్తన చెందడానికి పట్టే వ్యవధిని సూచిస్తుంది. చిన్న హాప్ పొడవులు అంచనా వేయడానికి ఎక్కువ సమయం అవసరం, కానీ అధిక పిచ్ ఖచ్చితత్వాన్ని ఇస్తాయి.",
+ "Batch Size": "బ్యాచ్ పరిమాణం",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "దీన్ని మీ జీపీయూలో అందుబాటులో ఉన్న వీఆర్ఏఎంతో అనుసంధానం చేసుకోవడం మంచిది. 4 యొక్క అమరిక మెరుగైన ఖచ్చితత్వాన్ని అందిస్తుంది కాని నెమ్మదిగా ప్రాసెసింగ్ చేస్తుంది, అయితే 8 వేగవంతమైన మరియు ప్రామాణిక ఫలితాలను అందిస్తుంది.",
+ "Save Every Epoch": "ప్రతి యుగాన్ని కాపాడండి",
+ "Determine at how many epochs the model will saved at.": "మోడల్ ఎన్ని యుగాలలో సేవ్ చేయబడుతుందో నిర్ణయించండి.",
+ "Total Epoch": "మొత్తం యుగం",
+ "Specifies the overall quantity of epochs for the model training process.": "మోడల్ ట్రైనింగ్ ప్రాసెస్ కొరకు యుగాల యొక్క మొత్తం పరిమాణాన్ని నిర్దేశిస్తుంది.",
+ "Pretrained": "ప్రీ ట్రైనింగ్ చేయబడింది",
+ "Save Only Latest": "సేవ్ ఓన్లీ లేటెస్ట్",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "ఈ సెట్టింగ్ ని ఎనేబుల్ చేయడం వల్ల G మరియు D ఫైల్స్ వాటి ఇటీవలి వెర్షన్ లను మాత్రమే సేవ్ చేస్తాయి, స్టోరేజీ స్పేస్ ని సమర్థవంతంగా సంరక్షిస్తాయి.",
+ "Save Every Weights": "ప్రతి బరువులను ఆదా చేయండి",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "ఈ అమరిక ప్రతి యుగం ముగింపులో మోడల్ యొక్క బరువులను సేవ్ చేయడానికి మిమ్మల్ని అనుమతిస్తుంది.",
+ "Custom Pretrained": "Custom Pretrained",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "కస్టమ్ ప్రీ-ట్రైన్డ్ మోడళ్లను ఉపయోగించడం మెరుగైన ఫలితాలకు దారితీస్తుంది, ఎందుకంటే నిర్దిష్ట ఉపయోగం కేసుకు అనుగుణంగా అత్యంత తగిన ప్రీట్రైనింగ్ మోడళ్లను ఎంచుకోవడం పనితీరును గణనీయంగా మెరుగుపరుస్తుంది.",
+ "Upload Pretrained Model": "ప్రీ ట్రైన్డ్ మోడల్ అప్ లోడ్ చేయండి",
+ "Refresh Custom Pretraineds": "రిఫ్రెష్ కస్టమ్ ప్రీ ట్రైన్డ్ లు",
+ "Pretrained Custom Settings": "ముందస్తుగా శిక్షణ పొందిన కస్టమ్ సెట్టింగ్ లు",
+ "The file you dropped is not a valid pretrained file. Please try again.": "మీరు డ్రాప్ చేసిన ఫైల్ చెల్లుబాటు అయ్యే ముందస్తు శిక్షణ పొందిన ఫైల్ కాదు. దయచేసి మళ్లీ ప్రయత్నించండి.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "డ్రాప్ డౌన్ మెనూలో ముందుగా శిక్షణ పొందిన ఫైల్ ను చూడటం కొరకు రిఫ్రెష్ బటన్ మీద క్లిక్ చేయండి.",
+ "Pretrained G Path": "Custom Pretrained G",
+ "Pretrained D Path": "Custom Pretrained D",
+ "GPU Settings": "GPU సెట్టింగ్ లు",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "మెరుగైన GPU ఆర్కిటెక్చర్ ఉన్న వినియోగదారుల కొరకు సిఫారసు చేయబడ్డ అధునాతన GPU సెట్టింగ్ లను సెట్ చేస్తుంది.",
+ "GPU Custom Settings": "GPU కస్టమ్ సెట్టింగ్ లు",
+ "GPU Number": "జిపియు నెంబరు",
+ "0 to ∞ separated by -": "0 నుండి ∞ వరకు దీని ద్వారా వేరు చేయబడతాయి -",
+ "GPU Information": "GPU సమాచారం",
+ "Pitch Guidance": "పిచ్ మార్గదర్శకత్వం",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "పిచ్ మార్గదర్శకత్వాన్ని ఉపయోగించడం ద్వారా, దాని పిచ్తో సహా అసలు స్వరం యొక్క స్వరాన్ని ప్రతిబింబించడం సాధ్యమవుతుంది. అసలు మెలోడీ లేదా పిచ్ నమూనాను సంరక్షించడం అవసరమైన గానం మరియు ఇతర సందర్భాలకు ఈ లక్షణం ముఖ్యంగా విలువైనది.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "మీ స్వంత శిక్షణ చేసేటప్పుడు ముందస్తు శిక్షణ పొందిన నమూనాలను ఉపయోగించండి. ఈ విధానం శిక్షణ వ్యవధిని తగ్గిస్తుంది మరియు మొత్తం నాణ్యతను పెంచుతుంది.",
+ "Extract Features": "ఎక్స్ ట్రాక్ట్ ఫీచర్లు",
+ "Start Training": "శిక్షణ ప్రారంభించండి",
+ "Generate Index": "ఇండెక్స్ జనరేట్ చేయండి",
+ "Voice Model": "వాయిస్ మోడల్",
+ "Select the voice model to use for the conversion.": "కన్వర్షన్ కొరకు ఉపయోగించాల్సిన వాయిస్ మోడల్ ని ఎంచుకోండి.",
+ "Index File": "ఇండెక్స్ ఫైల్",
+ "Select the index file to use for the conversion.": "మార్పిడి కొరకు ఉపయోగించాల్సిన ఇండెక్స్ ఫైల్ ని ఎంచుకోండి.",
+ "Refresh": "రిఫ్రెష్",
+ "Unload Voice": "వాయిస్ ను అన్ లోడ్ చేయండి",
+ "Single": "ఏక",
+ "Upload Audio": "ఆడియో అప్ లోడ్ చేయండి",
+ "Select Audio": "ఆడియోను ఎంచుకోండి",
+ "Select the audio to convert.": "కన్వర్ట్ చేయడానికి ఆడియోను ఎంచుకోండి.",
+ "Advanced Settings": "అధునాతన సెట్టింగ్ లు",
+ "Clear Outputs (Deletes all audios in assets/audios)": "క్లియర్ అవుట్ పుట్స్ (అసెట్స్/ఆడియోల్లోని అన్ని ఆడియోలను తొలగిస్తుంది)",
+ "Custom Output Path": "కస్టమ్ అవుట్ పుట్ మార్గం",
+ "Output Path": "అవుట్ పుట్ మార్గం",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "అవుట్ పుట్ ఆడియో సేవ్ చేయబడే మార్గం, ఆస్తులు/ఆడియోలు/output.wav",
+ "Split Audio": "స్ప్లిట్ ఆడియో",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "కొన్ని సందర్భాల్లో మెరుగైన ఫలితాలను పొందడానికి ఆడియోను భాగాలుగా విభజించండి.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "పాటల మార్పిడి కోసం సిఫార్సు చేయబడిన మీ అంచనాలకు మృదువైన ఆటోట్యూన్ను వర్తించండి.",
+ "Clean Audio": "క్లీన్ ఆడియో",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "ఆడియోలు మాట్లాడటానికి సిఫారసు చేయబడిన నాయిస్ డిటెక్షన్ అల్గారిథమ్ లను ఉపయోగించి మీ ఆడియో అవుట్ పుట్ ను శుభ్రం చేయండి.",
+ "Clean Strength": "క్లీన్ స్ట్రెంత్",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "మీకు కావలసిన ఆడియోకు క్లీన్-అప్ స్థాయిని సెట్ చేయండి, మీరు దానిని ఎంత పెంచితే అది మరింత క్లీన్ అవుతుంది, కానీ ఆడియో మరింత కంప్రెస్ అయ్యే అవకాశం ఉంది.",
+ "Pitch": "గొంతు",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ఆడియో యొక్క పిచ్ సెట్ చేయండి, విలువ ఎక్కువ, పిచ్ ఎక్కువ.",
+ "Filter Radius": "Filter వ్యాసార్థం",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "సంఖ్య మూడు కంటే ఎక్కువగా లేదా సమానంగా ఉంటే, సేకరించిన టోన్ ఫలితాలపై మధ్యస్థ వడపోతను ఉపయోగించడం శ్వాసక్రియను తగ్గించే సామర్థ్యాన్ని కలిగి ఉంటుంది.",
+ "Search Feature Ratio": "శోధన ఫీచర్ నిష్పత్తి",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "ఇండెక్స్ ఫైలు ప్రభావం; అధిక విలువ ఎక్కువ ప్రభావానికి అనుగుణంగా ఉంటుంది. అయినప్పటికీ, తక్కువ విలువలను ఎంచుకోవడం ఆడియోలో ఉన్న కళాఖండాలను తగ్గించడానికి సహాయపడుతుంది.",
+ "Volume Envelope": "వాల్యూమ్ కవరు",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "అవుట్ పుట్ యొక్క వాల్యూమ్ కవర్ తో ప్రత్యామ్నాయం చేయండి లేదా కలపండి. నిష్పత్తి 1 కి దగ్గరగా ఉంటే, అవుట్ పుట్ కవరు ఎక్కువగా ఉపయోగించబడుతుంది.",
+ "Protect Voiceless Consonants": "వాయిస్ లెస్ కన్సొనెంట్ లను సంరక్షించండి",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "ఎలక్ట్రో-అకౌస్టిక్ చిరిగిపోవడం మరియు ఇతర కళాఖండాలను నిరోధించడానికి విభిన్న ధ్వనులు మరియు శ్వాస శబ్దాలను సంరక్షించండి. పరామీటర్ ను దాని గరిష్ట విలువ 0.5కు లాగడం సమగ్ర రక్షణను అందిస్తుంది. ఏదేమైనా, ఈ విలువను తగ్గించడం వల్ల ఇండెక్సింగ్ ప్రభావాన్ని తగ్గించేటప్పుడు రక్షణ పరిధి తగ్గుతుంది.",
+ "Pitch extraction algorithm": "పిచ్ వెలికితీత అల్గోరిథం",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "ఆడియో మార్పిడి కోసం ఉపయోగించే పిచ్ వెలికితీత అల్గోరిథం. డిఫాల్ట్ అల్గోరిథం ఆర్ఎమ్విపిఇ, ఇది చాలా సందర్భాలలో సిఫార్సు చేయబడింది.",
+ "Convert": "మార్చండి",
+ "Export Audio": "Export Audio",
+ "Batch": "దొంతర",
+ "Input Folder": "ఇన్ పుట్ ఫోల్డర్",
+ "Select the folder containing the audios to convert.": "కన్వర్ట్ చేయడానికి ఆడియోలను కలిగి ఉన్న సంచికను ఎంచుకోండి.",
+ "Enter input path": "ఇన్ పుట్ మార్గాన్ని నమోదు చేయండి",
+ "Output Folder": "అవుట్ పుట్ ఫోల్డర్",
+ "Select the folder where the output audios will be saved.": "అవుట్ పుట్ ఆడియోలు సేవ్ చేయబడే ఫోల్డర్ ను ఎంచుకోండి.",
+ "Enter output path": "అవుట్ పుట్ మార్గాన్ని నమోదు చేయండి",
+ "Get information about the audio": "ఆడియో గురించి సమాచారం పొందండి",
+ "Information about the audio file": "ఆడియో ఫైలు గురించి సమాచారం",
+ "Waiting for information...": "సమాచారం కోసం ఎదురుచూస్తూ...",
+ "## Voice Blender": "## వాయిస్ బ్లెండర్",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "రెండు వాయిస్ మోడళ్లను ఎంచుకోండి, మీకు కావలసిన మిశ్రమ శాతాన్ని సెట్ చేయండి మరియు వాటిని పూర్తిగా కొత్త గొంతులో కలపండి.",
+ "Voice Blender": "వాయిస్ బ్లెండర్",
+ "Drag and drop your model here": "మీ మోడల్ ని ఇక్కడ డ్రాగ్ చేయండి మరియు డ్రాప్ చేయండి",
+ "You can also use a custom path.": "మీరు కస్టమ్ మార్గాన్ని కూడా ఉపయోగించవచ్చు.",
+ "Blend Ratio": "మిశ్రమ నిష్పత్తి",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "పొజిషన్ ను ఒక వైపు లేదా మరొక వైపుకు మరింత సర్దుబాటు చేయడం వల్ల మోడల్ మొదటి లేదా రెండవ వైపు మరింత పోలి ఉంటుంది.",
+ "Fusion": "ఫ్యూజన్",
+ "Path to Model": "మోడల్ కు మార్గం[మార్చు]",
+ "Enter path to model": "మోడల్ కు మార్గాన్ని నమోదు చేయండి",
+ "Model information to be placed": "మోడల్ సమాచారం ఉంచాలి",
+ "Inroduce the model information": "మోడల్ సమాచారాన్ని ఇన్ డ్యూక్ చేయండి",
+ "The information to be placed in the model (You can leave it blank or put anything).": "మోడల్ లో ఉంచాల్సిన సమాచారం (మీరు దానిని ఖాళీగా ఉంచవచ్చు లేదా ఏదైనా ఉంచవచ్చు).",
+ "View model information": "నమూనా సమాచారాన్ని వీక్షించండి",
+ "Introduce the model pth path": "మోడల్ పిటిహెచ్ మార్గాన్ని పరిచయం చేయండి",
+ "View": "దృశ్యం",
+ "Model extraction": "నమూనా వెలికితీత",
+ "Model conversion": "నమూనా మార్పిడి",
+ "Pth file": "PTH ఫైల్",
+ "Output of the pth file": "పిటిహెచ్ ఫైల్ యొక్క అవుట్ పుట్",
+ "# How to Report an Issue on GitHub": "# గిట్హబ్లో సమస్యను ఎలా రిపోర్ట్ చేయాలి",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. మీరు ఎదుర్కొంటున్న సమస్యను రికార్డ్ చేయడం ప్రారంభించడానికి దిగువ 'రికార్డ్ స్క్రీన్' బటన్పై క్లిక్ చేయండి.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. మీరు సమస్యను రికార్డ్ చేయడం పూర్తి చేసిన తర్వాత, 'స్టాప్ రికార్డింగ్' బటన్పై క్లిక్ చేయండి (అదే బటన్, కానీ మీరు చురుకుగా రికార్డ్ చేస్తున్నారా లేదా అనే దానిపై ఆధారపడి లేబుల్ మారుతుంది).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. 'గిట్హబ్ ఇష్యూస్'(https://github.com/IAHispano/Applio/issues)లోకి వెళ్లి 'న్యూ ఇష్యూ' బటన్పై క్లిక్ చేయాలి.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. ఇచ్చిన ఇష్యూ టెంప్లేట్ ను పూర్తి చేసి, అవసరమైన విధంగా వివరాలను పొందుపర్చేలా చూసుకోవాలి మరియు మునుపటి దశ నుండి రికార్డ్ చేయబడిన ఫైల్ ను అప్ లోడ్ చేయడానికి ఆస్తుల విభాగాన్ని ఉపయోగించండి.",
+ "Record Screen": "స్క్రీన్ రికార్డ్ చేయండి",
+ "Record": "నమోదు",
+ "Stop Recording": "రికార్డింగ్ ఆపండి",
+ "Introduce the model .pth path": "మోడల్ .pth మార్గాన్ని పరిచయం చేయండి",
+ "See Model Information": "మోడల్ సమాచారం చూడండి",
+ "## Download Model": "## డౌన్ లోడ్ మోడల్",
+ "Model Link": "మోడల్ లింక్",
+ "Introduce the model link": "మోడల్ లింక్ ను పరిచయం చేయండి",
+ "Download Model": "మోడల్ డౌన్ లోడ్ చేయండి",
+ "## Drop files": "## డ్రాప్ ఫైళ్లు",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "మీ .pth ఫైల్ మరియు .index ఫైల్ ని ఈ స్పేస్ లోకి లాగండి. ఒకదాన్ని లాగి, తర్వాత మరొకటి లాగండి.",
+ "TTS Voices": "టిటిఎస్ వాయిస్",
+ "Select the TTS voice to use for the conversion.": "కన్వర్షన్ కొరకు ఉపయోగించడానికి TTS వాయిస్ ని ఎంచుకోండి.",
+ "Text to Synthesize": "సంశ్లేషణ చేయడానికి టెక్స్ట్",
+ "Enter the text to synthesize.": "సంశ్లేషణ చేయడానికి టెక్స్ట్ ను నమోదు చేయండి.",
+ "Or you can upload a .txt file": "లేదా .txt ఫైల్ ను అప్ లోడ్ చేయవచ్చు.",
+ "Enter text to synthesize": "సంశ్లేషణ చేయడానికి టెక్స్ట్ ను నమోదు చేయండి",
+ "Output Path for TTS Audio": "TTS ఆడియో కొరకు అవుట్ పుట్ మార్గం",
+ "Output Path for RVC Audio": "RVC ఆడియో కొరకు అవుట్ పుట్ మార్గం",
+ "Enable Applio integration with Discord presence": "Discord సమక్షంలో Applio ఇంటిగ్రేషన్ ప్రారంభించు",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "ఇది డిస్కార్డ్ లో ప్రస్తుత అప్లైయో యాక్టివిటీని ప్రదర్శించే అవకాశాన్ని యాక్టివేట్ చేస్తుంది.",
+ "Enable Applio integration with applio.org/models using flask": "ఫ్లాస్క్ ను ఉపయోగించి applio.org/models అప్లికేషన్ ఇంటిగ్రేషన్ ప్రారంభించండి",
+ "It will activate the possibility of downloading models with a click from the website.": "వెబ్ సైట్ నుంచి క్లిక్ తో మోడళ్లను డౌన్ లోడ్ చేసుకునే సదుపాయాన్ని యాక్టివేట్ చేస్తుంది.",
+ "Theme": "థీమ్",
+ "Select the theme you want to use. (Requires restarting Applio)": "మీరు ఉపయోగించాలనుకుంటున్న థీమ్ ఎంచుకోండి. (అప్లికేషన్ ను పునఃప్రారంభించడం అవసరం)",
+ "Language": "భాష",
+ "Select the language you want to use. (Requires restarting Applio)": "మీరు ఉపయోగించాలనుకుంటున్న భాషను ఎంచుకోండి. (అప్లికేషన్ ను పునఃప్రారంభించడం అవసరం)",
+ "Plugin Installer": "Plugin Installer",
+ "Drag your plugin.zip to install it": "దీన్ని ఇన్ స్టాల్ చేయడానికి మీ plugin.zip లాగండి",
+ "Version Checker": "Version Checker",
+ "Check which version of Applio is the latest to see if you need to update.": "అప్లియో యొక్క ఏ వెర్షన్ లేటెస్ట్ అని చెక్ చేయండి, మీరు అప్ డేట్ చేయాల్సిన అవసరం ఉందా అని చూడండి.",
+ "Check for updates": "అప్ డేట్ ల కొరకు చెక్ చేయండి"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/th_TH.json b/assets/i18n/languages/th_TH.json
new file mode 100644
index 0000000000000000000000000000000000000000..de1ade140e2f422e3ee9adf1b78403e98fec2f94
--- /dev/null
+++ b/assets/i18n/languages/th_TH.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "สุดยอดเครื่องมือโคลนเสียงที่ได้รับการปรับให้เหมาะสมอย่างพิถีพิถันเพื่อพลังที่ไม่มีใครเทียบได้ความเป็นโมดูลและประสบการณ์ที่ใช้งานง่าย",
+ "This section contains some extra utilities that often may be in experimental phases.": "ส่วนนี้ประกอบด้วยยูทิลิตี้พิเศษบางอย่างที่มักจะอยู่ในขั้นตอนการทดลอง",
+ "Output Information": "ข้อมูลเอาต์พุต",
+ "The output information will be displayed here.": "ข้อมูลผลลัพธ์จะแสดงที่นี่",
+ "Inference": "การอนุมาน",
+ "Train": "รถไฟ",
+ "Extra": "พิเศษ",
+ "Merge Audios": "รวมไฟล์เสียง",
+ "Processing": "ประมวล ผล",
+ "Audio Analyzer": "เครื่องวิเคราะห์เสียง",
+ "Model Information": "ข้อมูลรุ่น",
+ "Plugins": "ปลั๊กอิน",
+ "Download": "ดาวน์โหลด",
+ "Report a Bug": "รายงานข้อบกพร่อง",
+ "Settings": "การตั้งค่า",
+ "Preprocess": "กระบวนการล่วงหน้า",
+ "Model Name": "ชื่อรุ่น",
+ "Name of the new model.": "ชื่อของรุ่นใหม่",
+ "Enter model name": "ใส่ชื่อรุ่น",
+ "Dataset Path": "เส้นทางชุดข้อมูล",
+ "Path to the dataset folder.": "เส้นทางไปยังโฟลเดอร์ชุดข้อมูล",
+ "Refresh Datasets": "รีเฟรชชุดข้อมูล",
+ "Dataset Creator": "ผู้สร้างชุดข้อมูล",
+ "Dataset Name": "ชื่อชุดข้อมูล",
+ "Name of the new dataset.": "ชื่อของชุดข้อมูลใหม่",
+ "Enter dataset name": "ป้อนชื่อชุดข้อมูล",
+ "Upload Audio Dataset": "อัปโหลดชุดข้อมูลเสียง",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "เพิ่มไฟล์เสียงลงในชุดข้อมูลเรียบร้อยแล้ว โปรดคลิกปุ่มดําเนินการล่วงหน้า",
+ "Enter dataset path": "ป้อนเส้นทางชุดข้อมูล",
+ "Sampling Rate": "Samp อัตราลิง",
+ "The sampling rate of the audio files.": "อัตราการสุ่มตัวอย่างของไฟล์เสียง",
+ "RVC Version": "รุ่น RVC",
+ "The RVC version of the model.": "รุ่น RVC ของรุ่น",
+ "Preprocess Dataset": "ชุดข้อมูลประมวลผลล่วงหน้า",
+ "Extract": "สกัด",
+ "Hop Length": "ความยาวกระโดด",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "หมายถึงระยะเวลาที่ระบบใช้ในการเปลี่ยนไปใช้การเปลี่ยนแปลงระดับเสียงที่สําคัญ ความยาวฮ็อพที่เล็กกว่าต้องใช้เวลามากขึ้นในการอนุมาน แต่มักจะให้ความแม่นยําของพิทช์ที่สูงขึ้น",
+ "Batch Size": "ขนาดแบทช์",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "ขอแนะนําให้จัดตําแหน่งให้ตรงกับ VRAM ที่มีอยู่ของ GPU ของคุณ การตั้งค่า 4 ให้ความแม่นยําที่ดีขึ้น แต่การประมวลผลช้าลง ในขณะที่ 8 ให้ผลลัพธ์ที่เร็วขึ้นและเป็นมาตรฐาน",
+ "Save Every Epoch": "บันทึกทุกยุค",
+ "Determine at how many epochs the model will saved at.": "กําหนดจํานวนยุคที่โมเดลจะบันทึกไว้",
+ "Total Epoch": "ยุคทั้งหมด",
+ "Specifies the overall quantity of epochs for the model training process.": "ระบุปริมาณโดยรวมของยุคสําหรับกระบวนการฝึกอบรมแบบจําลอง",
+ "Pretrained": "ได้รับการฝึกฝนล่วงหน้า",
+ "Save Only Latest": "บันทึกเฉพาะล่าสุด",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "การเปิดใช้งานการตั้งค่านี้จะส่งผลให้ไฟล์ G และ D บันทึกเฉพาะเวอร์ชันล่าสุด ซึ่งช่วยประหยัดพื้นที่จัดเก็บได้อย่างมีประสิทธิภาพ",
+ "Save Every Weights": "ประหยัดทุกตุ้มน้ําหนัก",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "การตั้งค่านี้ช่วยให้คุณประหยัดน้ําหนักของโมเดลเมื่อสิ้นสุดแต่ละยุค",
+ "Custom Pretrained": "กําหนดเอง Pretrained",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "การใช้โมเดลที่ผ่านการฝึกอบรมล่วงหน้าแบบกําหนดเองสามารถนําไปสู่ผลลัพธ์ที่เหนือกว่า เนื่องจากการเลือกโมเดลที่ผ่านการฝึกอบรมล่วงหน้าที่เหมาะสมที่สุดซึ่งปรับให้เหมาะกับกรณีการใช้งานเฉพาะสามารถเพิ่มประสิทธิภาพได้อย่างมาก",
+ "Upload Pretrained Model": "อัปโหลดโมเดลที่ผ่านการฝึกอบรม",
+ "Refresh Custom Pretraineds": "รีเฟรช Custom Pretraineds",
+ "Pretrained Custom Settings": "การตั้งค่าแบบกําหนดเองที่ผ่านการฝึกอบรมล่วงหน้า",
+ "The file you dropped is not a valid pretrained file. Please try again.": "ไฟล์ที่คุณทิ้งไม่ใช่ไฟล์ที่ผ่านการฝึกอบรมล่วงหน้าที่ถูกต้อง โปรดลองอีกครั้ง",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "คลิกปุ่มรีเฟรชเพื่อดูไฟล์ที่ฝึกไว้ล่วงหน้าในเมนูแบบเลื่อนลง",
+ "Pretrained G Path": "Custom Pretrained G",
+ "Pretrained D Path": "กําหนดเอง Pretrained D",
+ "GPU Settings": "การตั้งค่า GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "ตั้งค่า GPU ขั้นสูง แนะนําสําหรับผู้ใช้ที่มีสถาปัตยกรรม GPU ที่ดีกว่า",
+ "GPU Custom Settings": "การตั้งค่าแบบกําหนดเองของ GPU",
+ "GPU Number": "หมายเลข GPU",
+ "0 to ∞ separated by -": "0 ถึง ∞ คั่นด้วย -",
+ "GPU Information": "ข้อมูล GPU",
+ "Pitch Guidance": "คําแนะนําสนาม",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "ด้วยการใช้คําแนะนําระดับเสียง จึงเป็นไปได้ที่จะสะท้อนน้ําเสียงของเสียงต้นฉบับ รวมถึงระดับเสียงด้วย คุณลักษณะนี้มีประโยชน์อย่างยิ่งสําหรับการร้องเพลงและสถานการณ์อื่นๆ ที่การรักษาท่วงทํานองดั้งเดิมหรือรูปแบบระดับเสียงเป็นสิ่งสําคัญ",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "ใช้โมเดลที่ผ่านการฝึกอบรมล่วงหน้าเมื่อฝึกของคุณเอง วิธีการนี้ช่วยลดระยะเวลาการฝึกอบรมและปรับปรุงคุณภาพโดยรวม",
+ "Extract Features": "คุณสมบัติสารสกัด",
+ "Start Training": "เริ่มการฝึกอบรม",
+ "Generate Index": "สร้างดัชนี",
+ "Voice Model": "รูปแบบเสียง",
+ "Select the voice model to use for the conversion.": "เลือกรูปแบบเสียงที่จะใช้สําหรับการแปลง",
+ "Index File": "ไฟล์ดัชนี",
+ "Select the index file to use for the conversion.": "เลือกไฟล์ดัชนีที่จะใช้สําหรับการแปลง",
+ "Refresh": "ฟื้นฟู",
+ "Unload Voice": "ยกเลิกการโหลดเสียง",
+ "Single": "โสด",
+ "Upload Audio": "อัพโหลดเสียง",
+ "Select Audio": "เลือกเสียง",
+ "Select the audio to convert.": "เลือกเสียงที่จะแปลง",
+ "Advanced Settings": "การตั้งค่าขั้นสูง",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Clear Outputs (ลบไฟล์เสียงทั้งหมดในเนื้อหา/ไฟล์เสียง)",
+ "Custom Output Path": "เส้นทางเอาต์พุตที่กําหนดเอง",
+ "Output Path": "เส้นทางเอาต์พุต",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "เส้นทางที่จะบันทึกเสียงเอาต์พุตโดยค่าเริ่มต้นในเนื้อหา/ไฟล์เสียง/output.wav",
+ "Split Audio": "แยกเสียง",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "แบ่งเสียงออกเป็นชิ้นๆ เพื่อการอนุมานเพื่อให้ได้ผลลัพธ์ที่ดีขึ้นในบางกรณี",
+ "Autotune": "ออโต้จูน",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "ใช้ซอฟต์ออโต้จูนกับการอนุมานของคุณ ซึ่งแนะนําสําหรับการแปลงการร้องเพลง",
+ "Clean Audio": "ทําความสะอาดเสียง",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "ทําความสะอาดเอาต์พุตเสียงของคุณโดยใช้อัลกอริธึมการตรวจจับเสียงรบกวน ซึ่งแนะนําสําหรับการพูดไฟล์เสียง",
+ "Clean Strength": "ความแข็งแรงสะอาด",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "ตั้งค่าระดับการล้างข้อมูลเป็นเสียงที่คุณต้องการยิ่งคุณเพิ่มมากเท่าไหร่ก็ยิ่งทําความสะอาดได้มากขึ้นเท่านั้น แต่เป็นไปได้ว่าเสียงจะถูกบีบอัดมากขึ้น",
+ "Pitch": "ขว้าง",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "ตั้งค่าระดับเสียง ยิ่งค่าสูง ระดับเสียงก็จะยิ่งสูงขึ้น",
+ "Filter Radius": "รัศมีตัวกรอง",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "หากจํานวนมากกว่าหรือเท่ากับสามการใช้การกรองค่ามัธยฐานกับผลลัพธ์ของโทนเสียงที่รวบรวมได้มีศักยภาพในการลดการหายใจ",
+ "Search Feature Ratio": "อัตราส่วนคุณลักษณะการค้นหา",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "อิทธิพลที่เกิดจากไฟล์ดัชนี ค่าที่สูงขึ้นสอดคล้องกับอิทธิพลที่มากขึ้น อย่างไรก็ตาม การเลือกใช้ค่าที่ต่ํากว่าสามารถช่วยบรรเทาสิ่งประดิษฐ์ที่มีอยู่ในเสียงได้",
+ "Volume Envelope": "ซองจดหมายปริมาณ",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "แทนที่หรือผสมกับซองปริมาตรของเอาต์พุต ยิ่งอัตราส่วนใกล้เคียงกับ 1 มากเท่าไหร่ก็ยิ่งใช้ซองเอาต์พุตมากขึ้นเท่านั้น",
+ "Protect Voiceless Consonants": "ปกป้องพยัญชนะไร้เสียง",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "ปกป้องพยัญชนะและเสียงหายใจที่แตกต่างกันเพื่อป้องกันการฉีกขาดของอิเล็กโทรอะคูสติกและสิ่งประดิษฐ์อื่นๆ การดึงพารามิเตอร์ไปที่ค่าสูงสุด 0.5 ให้การป้องกันที่ครอบคลุม อย่างไรก็ตาม การลดค่านี้อาจลดขอบเขตของการป้องกันในขณะที่อาจบรรเทาผลกระทบจากการจัดทําดัชนี",
+ "Pitch extraction algorithm": "อัลกอริธึมการแยกระดับเสียง",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "อัลกอริธึมการแยกระดับเสียงเพื่อใช้สําหรับการแปลงเสียง อัลกอริทึมเริ่มต้นคือ rmvpe ซึ่งแนะนําสําหรับกรณีส่วนใหญ่",
+ "Convert": "แปลง",
+ "Export Audio": "ส่งออกเสียง",
+ "Batch": "ชุด",
+ "Input Folder": "โฟลเดอร์อินพุต",
+ "Select the folder containing the audios to convert.": "เลือกโฟลเดอร์ที่มีไฟล์เสียงที่จะแปลง",
+ "Enter input path": "ป้อนเส้นทางอินพุต",
+ "Output Folder": "โฟลเดอร์ผลลัพธ์",
+ "Select the folder where the output audios will be saved.": "เลือกโฟลเดอร์ที่จะบันทึกไฟล์เสียงที่ส่งออก",
+ "Enter output path": "ป้อนเส้นทางเอาต์พุต",
+ "Get information about the audio": "รับข้อมูลเกี่ยวกับเสียง",
+ "Information about the audio file": "ข้อมูลเกี่ยวกับไฟล์เสียง",
+ "Waiting for information...": "รอข้อมูล...",
+ "## Voice Blender": "## เครื่องปั่นเสียง",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "เลือกรุ่นเสียงสองแบบตั้งค่าเปอร์เซ็นต์การผสมที่คุณต้องการและผสมผสานเป็นเสียงใหม่ทั้งหมด",
+ "Voice Blender": "เครื่องปั่นเสียง",
+ "Drag and drop your model here": "ลากและวางโมเดลของคุณที่นี่",
+ "You can also use a custom path.": "คุณยังสามารถใช้เส้นทางที่กําหนดเอง",
+ "Blend Ratio": "อัตราส่วนการผสม",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "การปรับตําแหน่งไปทางด้านใดด้านหนึ่งมากขึ้นจะทําให้โมเดลคล้ายกับครั้งแรกหรือครั้งที่สองมากขึ้น",
+ "Fusion": "ฟิวชั่น",
+ "Path to Model": "เส้นทางสู่โมเดล",
+ "Enter path to model": "ป้อนเส้นทางไปยังแบบจําลอง",
+ "Model information to be placed": "ข้อมูลรุ่นที่จะวาง",
+ "Inroduce the model information": "ให้ข้อมูลโมเดล",
+ "The information to be placed in the model (You can leave it blank or put anything).": "ข้อมูลที่จะวางในแบบจําลอง (คุณสามารถเว้นว่างไว้หรือใส่อะไรก็ได้)",
+ "View model information": "ดูข้อมูลรุ่น",
+ "Introduce the model pth path": "แนะนําเส้นทาง pth ของโมเดล",
+ "View": "ทิวทัศน์",
+ "Model extraction": "การสกัดแบบจําลอง",
+ "Model conversion": "การแปลงโมเดล",
+ "Pth file": "ไฟล์ Pth",
+ "Output of the pth file": "เอาต์พุตของไฟล์ pth",
+ "# How to Report an Issue on GitHub": "# วิธีรายงานปัญหาใน GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. คลิกที่ปุ่ม 'บันทึกหน้าจอ' ด้านล่างเพื่อเริ่มบันทึกปัญหาที่คุณพบ",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. เมื่อคุณบันทึกปัญหาเสร็จแล้ว ให้คลิกที่ปุ่ม 'หยุดการบันทึก' (ปุ่มเดิม แต่ป้ายกํากับจะเปลี่ยนขึ้นอยู่กับว่าคุณกําลังบันทึกอยู่หรือไม่)",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. ไปที่ [GitHub Issues](https://github.com/IAHispano/Applio/issues) และคลิกที่ปุ่ม 'New Issue'",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. กรอกเทมเพลตปัญหาที่ให้มา เพื่อให้แน่ใจว่ามีรายละเอียดตามต้องการ และใช้ส่วนเนื้อหาเพื่ออัปโหลดไฟล์ที่บันทึกไว้จากขั้นตอนก่อนหน้า",
+ "Record Screen": "หน้าจอบันทึก",
+ "Record": "บันทึก",
+ "Stop Recording": "หยุดการบันทึก",
+ "Introduce the model .pth path": "แนะนําโมเดล .pth path",
+ "See Model Information": "ดูข้อมูลรุ่น",
+ "## Download Model": "## ดาวน์โหลดโมเดล",
+ "Model Link": "ลิงค์โมเดล",
+ "Introduce the model link": "แนะนําลิงค์โมเดล",
+ "Download Model": "ดาวน์โหลดโมเดล",
+ "## Drop files": "## วางไฟล์",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "ลากไฟล์ .pth และไฟล์ .index ลงในช่องว่างนี้ ลากหนึ่งแล้วลากอีกอันหนึ่ง",
+ "TTS Voices": "เสียง TTS",
+ "Select the TTS voice to use for the conversion.": "เลือกเสียง TTS ที่จะใช้สําหรับการแปลง",
+ "Text to Synthesize": "ข้อความที่จะสังเคราะห์",
+ "Enter the text to synthesize.": "ป้อนข้อความที่จะสังเคราะห์",
+ "Or you can upload a .txt file": "หรือคุณสามารถอัปโหลดไฟล์ .txt",
+ "Enter text to synthesize": "ป้อนข้อความที่จะสังเคราะห์",
+ "Output Path for TTS Audio": "เส้นทางเอาต์พุตสําหรับเสียง TTS",
+ "Output Path for RVC Audio": "เส้นทางเอาต์พุตสําหรับเสียง RVC",
+ "Enable Applio integration with Discord presence": "เปิดใช้งานการรวม Applio กับการแสดงตนของ Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "มันจะเปิดใช้งานความเป็นไปได้ในการแสดงกิจกรรม Applio ปัจจุบันใน Discord",
+ "Enable Applio integration with applio.org/models using flask": "เปิดใช้งานการรวม Applio กับ applio.org/models โดยใช้ขวด",
+ "It will activate the possibility of downloading models with a click from the website.": "มันจะเปิดใช้งานความเป็นไปได้ในการดาวน์โหลดโมเดลด้วยการคลิกจากเว็บไซต์",
+ "Theme": "หัวข้อ",
+ "Select the theme you want to use. (Requires restarting Applio)": "เลือกธีมที่คุณต้องการใช้ (ต้องรีสตาร์ท Applio)",
+ "Language": "ภาษา",
+ "Select the language you want to use. (Requires restarting Applio)": "เลือกภาษาที่คุณต้องการใช้ (ต้องรีสตาร์ท Applio)",
+ "Plugin Installer": "ตัวติดตั้งปลั๊กอิน",
+ "Drag your plugin.zip to install it": "ลาก plugin.zip ของคุณเพื่อติดตั้ง",
+ "Version Checker": "ตัวตรวจสอบเวอร์ชัน",
+ "Check which version of Applio is the latest to see if you need to update.": "ตรวจสอบว่า Applio เวอร์ชันใดเป็นเวอร์ชันล่าสุดเพื่อดูว่าคุณจําเป็นต้องอัปเดตหรือไม่",
+ "Check for updates": "ตรวจหาการอัปเดต"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/tr_TR.json b/assets/i18n/languages/tr_TR.json
new file mode 100644
index 0000000000000000000000000000000000000000..d5c6a81bccea22acd9e2f3887bce70a9033fe892
--- /dev/null
+++ b/assets/i18n/languages/tr_TR.json
@@ -0,0 +1,215 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Üstün ses kopyalama aracı, rakipsiz güç, modülerlik ve kullanıcı dostu deneyim için özenle optimize edilmiştir.\n[Türkçe çeviri: Enes](https://discord.com/users/1140031358006202468)",
+ "This section contains some extra utilities that often may be in experimental phases.": "Bu bölümde, genellikle deneysel aşamalarda olabilecek bazı ekstra yardımcı programlar bulunur.",
+ "Output Information": "Çıktı Bilgileri",
+ "The output information will be displayed here.": "Çıktı bilgileri burada görüntülenecektir.",
+
+ "Inference": "Arayüz",
+ "Train": "Eğitim",
+ "Extra": "Ekstra",
+ "Merge Audios": "Sesleri Birleştir",
+ "Processing": "İşleme",
+ "Audio Analyzer": "Ses Analizörü",
+ "Model Information": "Model Bilgileri",
+ "Plugins": "Eklentiler",
+ "Download": "İndir",
+ "Report a Bug": "Bir Hata Bildir",
+ "Settings": "Ayarlar",
+
+ "Preprocess": "İşlem Öncesi",
+ "Model Name": "Model Adı",
+ "Name of the new model.": "Yeni modelin adı.",
+ "Enter model name": "Model adını girin",
+ "Dataset Path": "Veri Kümesi Yolu",
+ "Path to the dataset folder.": "Veri kümesi klasörünün yolu.",
+ "Refresh Datasets": "Veri Kümelerini Yenile",
+ "Dataset Creator": "Veri Kümesi Oluşturucu",
+ "Dataset Name": "Veri Kümesi Adı",
+ "Name of the new dataset.": "Yeni veri kümesinin adı.",
+ "Enter dataset name": "Veri kümesi adını girin",
+ "Upload Audio Dataset": "Ses Veri Kümesini Yükle",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Ses dosyası veri kümesine başarıyla eklendi. Lütfen işlem öncesi düğmesine tıklayın.",
+ "Enter dataset path": "Veri kümesi yolunu girin",
+ "Sampling Rate": "Örnekleme Hızı",
+ "The sampling rate of the audio files.": "Ses dosyalarının örnekleme hızı.",
+ "RVC Version": "RVC Sürümü",
+ "The RVC version of the model.": "Modelin RVC sürümü.",
+ "Preprocess Dataset": "Veri Kümesini İşlem Öncesinden Geçir",
+
+ "Embedder Model": "Gömme Modeli",
+ "Model used for learning speaker embedding.": "Konuşmacı gömmelerini öğrenmek için kullanılan model.",
+ "Extract": "Çıkar",
+ "Hop Length": "Adım Uzunluğu",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Sistemin önemli bir perde değişikliğine geçmesi için gereken süreyi belirtir. Daha küçük adım uzunlukları, çıkarsama için daha fazla zaman gerektirir ancak daha yüksek perde doğruluğu sağlar.",
+ "Batch Size": "Toplu Boyutu",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Bunu GPU'nuzun mevcut VRAM'iyle uyumlu hale getirmeniz önerilir. 4'lük bir ayar daha iyi doğruluk ancak daha yavaş işleme sunarken, 8 daha hızlı ve standart sonuçlar sağlar.",
+ "Save Every Epoch": "Her Dönemde Kaydet",
+ "Determine at how many epochs the model will saved at.": "Modelin kaçıncı çağlarda kaydedileceğini belirleyin.",
+ "Total Epoch": "Toplam Çağ",
+ "Specifies the overall quantity of epochs for the model training process.": "Model eğitim süreci için gereken toplam dönem sayısını belirtir.",
+ "Pretrained": "Önceden Eğitimli",
+ "Save Only Latest": "Sadece Son Sürümü Kaydet",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Bu ayarın etkinleştirilmesi, G ve D dosyalarının yalnızca en son sürümlerinin kaydedilmesine neden olur ve depolama alanından tasarruf sağlar.",
+ "Save Every Weights": "Tüm Ağırlıkları Kaydet",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Bu ayar, her dönemin sonunda modelin ağırlıklarını kaydetmenizi sağlar.",
+ "Custom Pretrained": "Özel Önceden Eğitimli",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Özel önceden eğitilmiş modelleri kullanmak üstün sonuçlar doğurabilir, çünkü belirli kullanım durumuna göre uyarlanmış en uygun önceden eğitilmiş modelleri seçmek performansı önemli ölçüde artırabilir.",
+ "Upload Pretrained Model": "Önceden Eğitimli Model Yükle",
+ "Refresh Custom Pretraineds": "Özel Önceden Eğitimleri Yenile",
+ "Pretrained Custom Settings": "Önceden Eğitimli Özel Ayarlar",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Bıraktığınız dosya geçerli bir önceden eğitilmiş dosya değil. Lütfen tekrar deneyin.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Açılır menüde önceden eğitilmiş dosyayı görmek için yenile düğmesine tıklayın.",
+ "Pretrained G Path": "Özel Önceden Eğitimli G",
+ "Pretrained D Path": "Özel Önceden Eğitimli D",
+ "GPU Settings": "GPU Ayarları",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Gelişmiş GPU ayarlarını belirler, daha iyi GPU mimarisine sahip kullanıcılar için önerilir.",
+ "GPU Custom Settings": "GPU Özel Ayarları",
+ "GPU Number": "GPU Numarası",
+ "0 to ∞ separated by -": "0'dan sonsuza kadar - ile ayrılmış",
+ "GPU Information": "GPU Bilgileri",
+ "Pitch Guidance": "Perde Rehberliği",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Perde rehberliğini kullanarak, orijinal sesin tonlamasını, perde de dahil olmak üzere yansıtmak mümkün hale gelir. Bu özellik, özellikle orijinal melodiyi veya perde düzenini korumak için hayati önem taşıyan şarkı söyleme ve diğer senaryolar için değerlidir.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Kendi modelinizi eğitirken önceden eğitilmiş modelleri kullanın. Bu yaklaşım, eğitim süresini azaltır ve genel kaliteyi artırır.",
+ "Extract Features": "Özellikleri Çıkar",
+ "Overtraining Detector": "Aşırı Eğitim Dedektörü",
+ "Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data.": "Modelin eğitim verilerini çok iyi öğrenmesini ve yeni verilere genelleme yeteneğini kaybetmesini önlemek için aşırı eğitimi tespit edin.",
+ "Overtraining Detector Settings": "Aşırı Eğitim Dedektörü Ayarları",
+ "Overtraining Threshold": "Aşırı Eğitim Eşiği",
+ "Set the maximum number of epochs you want your model to stop training if no improvement is detected.": "Herhangi bir iyileşme tespit edilmezse modelinizin eğitimini durdurmasını istediğiniz maksimum epoch sayısını ayarlayın.",
+
+ "Start Training": "Eğitmeye Başla",
+ "Stop Training & Restart Applio": "Eğitimi Durdur ve Applio'yu Yeniden Başlat",
+ "Generate Index": "Dizin Oluştur",
+
+ "Export Model": "Modeli Dışa Aktar",
+ "The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive.": "'Yükle' düğmesi yalnızca google colab içindir: Dışa aktarılan dosyaları Google Drive'ınızdaki ApplioExported klasörüne yükler.",
+ "Exported Pth file": "Dışa Aktarılan Pth dosyası",
+ "Exported Index file": "Dışa Aktarılan Dizin dosyası",
+ "Select the pth file to be exported": "Dışa aktarılacak pth dosyasını seçin",
+ "Select the index file to be exported": "Dışa aktarılacak dizin dosyasını seçin",
+ "Upload": "Yükle",
+
+ "Voice Model": "Ses Modeli",
+ "Select the voice model to use for the conversion.": "Dönüştürme için kullanılacak ses modelini seçin.",
+ "Index File": "Dizin Dosyası",
+ "Select the index file to use for the conversion.": "Dönüştürme için kullanılacak dizin dosyasını seçin.",
+ "Refresh": "Yenile",
+ "Unload Voice": "Sesi Kaldır",
+
+ "Single": "Tek",
+ "Upload Audio": "Ses Yükle",
+ "Select Audio": "Ses Seç",
+ "Select the audio to convert.": "Dönüştürülecek sesi seçin.",
+ "Advanced Settings": "Gelişmiş Ayarlar",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Çıktıları Temizle (assets/audios klasöründeki tüm sesleri siler)",
+ "Custom Output Path": "Özel Çıktı Yolu",
+ "Output Path": "Çıktı Yolu",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Varsayılan olarak assets/audios/output.wav'da kaydedilecek çıktı sesinin yolu",
+ "Split Audio": "Sesi Böl",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Daha iyi sonuçlar elde etmek için bazı durumlarda çıkarım için sesi parçalara bölün.",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Çıkarımlarınıza yumuşak bir autotune uygulayın, şarkı dönüştürmeleri için önerilir.",
+ "Clean Audio": "Sesi Temizle",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Ses algılama algoritmaları kullanarak ses çıktınızı temizleyin, konuşan sesler için önerilir.",
+ "Clean Strength": "Temiz Gücü",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "İstediğiniz sese temizleme seviyesini ayarlayın, ne kadar artırırsanız o kadar temizler, ancak sesin daha fazla sıkıştırılması mümkündür.",
+ "Pitch": "Perde",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Sesin perdesini ayarlayın, değer ne kadar yüksek olursa perde o kadar yüksek olur.",
+ "Filter Radius": "Filtre Yarıçapı",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Sayı üçten büyük veya ona eşitse, toplanan ton sonuçlarına ortanca filtreleme uygulamak solunumu azaltabilir.",
+ "Search Feature Ratio": "Arama Özelliği Oranı",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Dizin dosyası tarafından uygulanan etki; daha yüksek bir değer daha büyük etkiye karşılık gelir. Ancak daha düşük değerler seçmek, sesteki var olan eserlerin azaltılmasına yardımcı olabilir.",
+ "Volume Envelope": "Ses Zarıfı",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Çıktının ses zarfı ile değiştirme veya karıştırma yapın. Oran 1'e ne kadar yakınsa, çıktı zarfı o kadar fazla kullanılır.",
+ "Protect Voiceless Consonants": "Sessiz Ünsüzleri Koru",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Elektro-akustik yırtılmayı ve diğer eserleri önlemek için farklı ünsüzleri ve solunum seslerini koruyun. Parametreyi maksimum değeri 0,5'e çekmek kapsamlı koruma sağlar. Ancak bu değeri azaltmak, koruma kapsamını azaltırken dizin oluşturma etkisini azaltabilir.",
+ "Pitch extraction algorithm": "Perde çıkarma algoritması",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Ses dönüşümü için kullanılacak perde çıkarma algoritması. Varsayılan algoritma, çoğu durum için önerilen rmvpedir.",
+
+ "Convert": "Dönüştür",
+ "Export Audio": "Sesi Dışa Aktar",
+
+ "Batch": "Toplu",
+ "Input Folder": "Giriş Klasörü",
+ "Select the folder containing the audios to convert.": "Dönüştürülecek sesleri içeren klasörü seçin.",
+ "Enter input path": "Giriş yolunu girin",
+ "Output Folder": "Çıktı Klasörü",
+ "Select the folder where the output audios will be saved.": "Çıktı seslerinin kaydedileceği klasörü seçin.",
+ "Enter output path": "Çıktı yolunu girin",
+
+ "Get information about the audio": "Ses hakkında bilgi al",
+
+ "## Voice Blender": "## Ses Karıştırıcı",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "İki ses modeli seçin, istediğiniz karışım yüzdesini ayarlayın ve bunları tamamen yeni bir sese karıştırın.",
+ "Voice Blender": "Ses Karıştırıcı",
+ "Drag and drop your model here": "Modelinizi buraya sürükleyip bırakın",
+ "You can also use a custom path.": "Özel bir yol da kullanabilirsiniz.",
+ "Blend Ratio": "Karışım Oranı",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Konumu bir yöne veya diğerine daha fazla ayarlayarak modeli birinci veya ikinciye daha benzer hale getireceksiniz.",
+ "Fusion": "Birleştirme",
+
+ "Path to Model": "Modele Giden Yol",
+ "Enter path to model": "Modele giden yolu girin",
+ "Model information to be placed": "Yerleştirilecek model bilgileri",
+ "Inroduce the model information": "Model bilgilerini tanıtın",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Modele yerleştirilecek bilgiler (Boş bırakabilir veya bir şey koyabilirsiniz).",
+ "View model information": "Model bilgilerini görüntüleyin",
+ "Introduce the model pth path": "Model pth yolunu tanıtın",
+ "View": "Görüntüle",
+ "Model extraction": "Model çıkarımı",
+ "Model conversion": "Model dönüştürme",
+ "Pth file": "Pth dosyası",
+ "Output of the pth file": "Pth dosyasının çıktısı",
+
+ "# How to Report an Issue on GitHub": "# GitHub'da Bir Sorunu Nasıl Bildirirsiniz",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Yaşadığınız sorunu kaydetmeye başlamak için aşağıdaki 'Ekran Kaydı' düğmesine tıklayın.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Sorunu kaydetmeyi tamamladığınızda, 'Kaydı Durdur' düğmesine tıklayın (aynı düğme, ancak aktif olarak kayıt yapıp yapmadığınıza bağlı olarak etiket değişir).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [GitHub Sorunlarına](https://github.com/IAHispano/Applio/issues) gidin ve 'Yeni Sorun' düğmesine tıklayın.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Sağlanan sorun şablonunu tamamlayın, gerektiği gibi ayrıntıları eklediğinizden emin olun ve önceki adımdan kaydedilen dosyayı yüklemek için varlıklar bölümünü kullanın.",
+
+ "Record Screen": "Ekranı Kaydet",
+ "Record": "Kayıt Yap",
+ "Stop Recording": "Kaydı Durdur",
+
+ "Introduce the model .pth path": "Model .pth yolunu tanıtın",
+ "See Model Information": "Model Bilgilerini Görün",
+
+ "## Download Model": "## Model İndirin",
+ "Model Link": "Model Bağlantısı",
+ "Introduce the model link": "Model bağlantısını tanıtın",
+ "Download Model": "Modeli İndirin",
+ "## Drop files": "## Dosyaları bırakın",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": ".pth dosyanızı ve .index dosyanızı bu alana sürükleyin. Önce birini sonra diğerini sürükleyin.",
+ "## Search Model": "## Modeli Ara",
+ "Search": "Ara",
+ "Introduce the model name to search.": "Aramak için model adını girin.",
+ "We couldn't find models by that name.": "Bu isim'e sahip bir model bulunamadı.",
+
+ "TTS Voices": "TTS Sesleri",
+ "Select the TTS voice to use for the conversion.": "Dönüşüm için kullanılacak TTS sesini seçin.",
+ "Text to Synthesize": "Sentezlemek İçin Metin",
+ "Enter the text to synthesize.": "Sentezlemek için metni girin.",
+ "Or you can upload a .txt file": "Veya bir .txt dosyası yükleyebilirsiniz",
+ "Enter text to synthesize": "Sentezlemek için metin girin",
+ "Output Path for TTS Audio": "TTS Sesinin Çıktı Yolu",
+ "Output Path for RVC Audio": "RVC Sesinin Çıktı Yolu",
+
+ "Enable Applio integration with Discord presence": "Discord oynuyor Applio entegrasyonunu etkinleştirin",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Applio geçerli etkinliğini Discord'da görüntüleme olasılığını etkinleştirir.",
+ "Enable Applio integration with applio.org/models using flask": "Flask kullanarak applio.org/models ile Applio entegrasyonunu etkinleştirin",
+ "It will activate the possibility of downloading models with a click from the website.": "Web sitesinden bir tıklama ile modellerin indirilmesi olasılığını etkinleştirir.",
+ "Enable fake GPU": "Sahte GPU'yu Etkinleştir",
+ "Training is currently unsupported due to the absence of a GPU. To activate the training tab, navigate to the settings tab and enable the 'Fake GPU' option.": "GPU olmadığı için eğitim şu anda desteklenmiyor. Eğitim sekmesini etkinleştirmek için ayarlar sekmesine gidin ve 'Sahte GPU' seçeneğini etkinleştirin.",
+ "Activates the train tab. However, please note that this device lacks GPU capabilities, hence training is not supported. This option is only for testing purposes. (This option will restart Applio)": "Eğitim sekmesini etkinleştirir. Ancak, bu cihazda GPU yetenekleri olmadığından, eğitim desteklenmez. Bu seçenek yalnızca test amaçlıdır. (Bu seçenek Applio'yu yeniden başlatacaktır)",
+ "Theme": "Tema",
+ "Select the theme you want to use. (Requires restarting Applio)": "Kullanmak istediğiniz temayı seçin. (Applio'yu yeniden başlatmayı gerektirir)",
+ "Language": "Dil",
+ "Select the language you want to use. (Requires restarting Applio)": "Kullanmak istediğiniz dili seçin. (Applio'yu yeniden başlatmayı gerektirir)",
+
+ "Plugin Installer": "Plugin Yükleyici",
+ "Drag your plugin.zip to install it": "Yüklemek için plugin.zip'inizi sürükleyin",
+
+ "Version Checker": "Sürüm Denetleyicisi",
+ "Check which version of Applio is the latest to see if you need to update.": "Güncellemeniz gerekip gerekmediğini görmek için Applio'nun hangi sürümünün en son olduğunu kontrol edin.",
+ "Check for updates": "Güncellemeleri kontrol et"
+}
diff --git a/assets/i18n/languages/uk_UK.json b/assets/i18n/languages/uk_UK.json
new file mode 100644
index 0000000000000000000000000000000000000000..8683eae280369468de7cf19b0e48358a0e856081
--- /dev/null
+++ b/assets/i18n/languages/uk_UK.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Найкращий інструмент для клонування голосу, ретельно оптимізований для неперевершеної потужності, модульності та зручного використання.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Цей розділ містить деякі додаткові утиліти, які часто можуть перебувати на експериментальних стадіях.",
+ "Output Information": "Вихідна інформація",
+ "The output information will be displayed here.": "Тут буде відображена вихідна інформація.",
+ "Inference": "Висновок",
+ "Train": "Поїзд",
+ "Extra": "Додаткові",
+ "Merge Audios": "Об'єднання аудіо",
+ "Processing": "Обробки",
+ "Audio Analyzer": "Аналізатор звуку",
+ "Model Information": "Інформація про модель",
+ "Plugins": "Плагіни",
+ "Download": "Завантажити",
+ "Report a Bug": "Повідомити про помилку",
+ "Settings": "Параметри",
+ "Preprocess": "Попередня обробка",
+ "Model Name": "Назва моделі",
+ "Name of the new model.": "Назва нової моделі.",
+ "Enter model name": "Введіть назву моделі",
+ "Dataset Path": "Шлях набору даних",
+ "Path to the dataset folder.": "Шлях до папки набору даних.",
+ "Refresh Datasets": "Оновити набори даних",
+ "Dataset Creator": "Творець набору даних",
+ "Dataset Name": "Ім'я набору даних",
+ "Name of the new dataset.": "Ім'я нового набору даних.",
+ "Enter dataset name": "Введіть назву набору даних",
+ "Upload Audio Dataset": "Завантажити аудіо набір даних",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Аудіофайл успішно додано до набору даних. Будь ласка, натисніть кнопку попередньої обробки.",
+ "Enter dataset path": "Введіть шлях набору даних",
+ "Sampling Rate": "Частота дискретизації",
+ "The sampling rate of the audio files.": "Частота дискретизації аудіофайлів.",
+ "RVC Version": "Версія RVC",
+ "The RVC version of the model.": "Версія моделі RVC.",
+ "Preprocess Dataset": "Набір даних попередньої обробки",
+ "Extract": "Екстракт",
+ "Hop Length": "Довжина хмелю",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Позначає час, необхідний для переходу системи до значної зміни висоти тону. Менша довжина стрибка вимагає більше часу для виведення, але, як правило, дає вищу точність кроку.",
+ "Batch Size": "Розмір партії",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Бажано вирівняти його з доступною відеопам'яттю вашого графічного процесора. Значення 4 забезпечує підвищену точність, але повільнішу обробку, тоді як 8 забезпечує швидші та стандартніші результати.",
+ "Save Every Epoch": "Врятуйте кожну епоху",
+ "Determine at how many epochs the model will saved at.": "Визначте, у скількох епохах збережеться модель.",
+ "Total Epoch": "Ціла епоха",
+ "Specifies the overall quantity of epochs for the model training process.": "Визначає загальну кількість епох для процесу навчання моделі.",
+ "Pretrained": "Попереднє навчання",
+ "Save Only Latest": "Зберігайте лише останні",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Увімкнення цього параметра призведе до того, що файли G і D зберігатимуть лише найновіші версії, ефективно заощаджуючи місце для зберігання.",
+ "Save Every Weights": "Збережіть кожну вагу",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Цей параметр дозволяє зберегти вагу моделі в кінці кожної епохи.",
+ "Custom Pretrained": "Індивідуальне попереднє навчання",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Використання спеціальних попередньо навчених моделей може призвести до чудових результатів, оскільки вибір найбільш підходящих попередньо навчених моделей, адаптованих до конкретного випадку використання, може значно підвищити продуктивність.",
+ "Upload Pretrained Model": "Завантажте попередньо навчену модель",
+ "Refresh Custom Pretraineds": "Оновлення користувацьких попередньо навчених",
+ "Pretrained Custom Settings": "Попередньо навчені користувацькі налаштування",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Файл, який ви скинули, не є дійсним попередньо навченим файлом. Будь ласка, спробуйте ще раз.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Натисніть кнопку «Оновити», щоб переглянути попередньо підготовлений файл у розкривному меню.",
+ "Pretrained G Path": "Спеціальна попередньо навчена G",
+ "Pretrained D Path": "Спеціальний попередньо навчений D",
+ "GPU Settings": "Налаштування графічного процесора",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Встановлює розширені налаштування графічного процесора, рекомендовані для користувачів із кращою архітектурою графічного процесора.",
+ "GPU Custom Settings": "Користувацькі налаштування графічного процесора",
+ "GPU Number": "Номер графічного процесора",
+ "0 to ∞ separated by -": "від 0 до ∞ розділені -",
+ "GPU Information": "Інформація про графічний процесор",
+ "Pitch Guidance": "Керівництво пітчем",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Використовуючи висоту тону, стає можливим відобразити інтонацію оригінального голосу, включаючи його висоту. Ця функція особливо цінна для співу та інших сценаріїв, де важливо зберегти оригінальну мелодію або висоту тону.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Використовуйте попередньо підготовлені моделі під час навчання власних. Такий підхід скорочує тривалість навчання і підвищує загальну якість.",
+ "Extract Features": "особливості витягу",
+ "Start Training": "Почати навчання",
+ "Generate Index": "Згенерувати індекс",
+ "Voice Model": "Голосова модель",
+ "Select the voice model to use for the conversion.": "Виберіть модель голосу, яку потрібно використовувати для перетворення.",
+ "Index File": "Індексний файл",
+ "Select the index file to use for the conversion.": "Виберіть файл покажчика, який буде використано для перетворення.",
+ "Refresh": "Оновити",
+ "Unload Voice": "Вивантажити голос",
+ "Single": "Одного",
+ "Upload Audio": "Завантажити аудіо",
+ "Select Audio": "Виберіть Аудіо",
+ "Select the audio to convert.": "Виберіть аудіо, яке потрібно конвертувати.",
+ "Advanced Settings": "Розширені налаштування",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Очистити виходи (видаляє всі аудіозаписи в ресурсах/аудіозаписах)",
+ "Custom Output Path": "Власний вихідний шлях",
+ "Output Path": "Вихідний шлях",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Шлях, за яким буде збережено вихідне аудіо, за замовчуванням у assets/audios/output.wav",
+ "Split Audio": "Розділене аудіо",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Розділіть аудіо на фрагменти для висновків, щоб отримати кращі результати в деяких випадках.",
+ "Autotune": "Автотюнінг",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Застосовуйте до своїх висновків м'яку автонастройку, рекомендовану для співочих перетворень.",
+ "Clean Audio": "Чистий звук",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Очистіть аудіовихід за допомогою алгоритмів виявлення шуму, рекомендованих для озвучування аудіо.",
+ "Clean Strength": "Чиста міцність",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Встановіть потрібний рівень очищення звуку, чим більше ви його збільшите, тим більше він буде очищений, але можливо, що звук буде більш стиснутим.",
+ "Pitch": "Крок",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Встановіть висоту звуку, чим вище значення, тим вище висота.",
+ "Filter Radius": "Радіус фільтра",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Якщо число більше або дорівнює трьом, використання медіанної фільтрації на результатах зібраного тону може призвести до зниження дихання.",
+ "Search Feature Ratio": "Співвідношення функцій пошуку",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Вплив, який чинить індексний файл; Більш високе значення відповідає більшому впливу. Однак вибір нижчих значень може допомогти пом'якшити артефакти, присутні в аудіо.",
+ "Volume Envelope": "Об'ємний конверт",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Замініть або змішайте з об'ємною огинаючою виводу. Чим ближче відношення до 1, тим більше використовується вихідна огинаюча.",
+ "Protect Voiceless Consonants": "Захист глухих приголосних",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Захистіть чіткі приголосні та дихальні звуки, щоб запобігти електроакустичним розривам та іншим артефактам. Підтягування параметра до максимального значення 0,5 забезпечує комплексний захист. Однак зменшення цього значення може зменшити ступінь захисту, потенційно пом'якшуючи ефект індексації.",
+ "Pitch extraction algorithm": "Алгоритм вилучення кроку",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Алгоритм вилучення висоти тону, який слід використовувати для перетворення звуку. Типовим алгоритмом є rmvpe, який рекомендується для більшості випадків.",
+ "Convert": "Перетворити",
+ "Export Audio": "Експорт аудіо",
+ "Batch": "Пакетний",
+ "Input Folder": "Папка введення",
+ "Select the folder containing the audios to convert.": "Виберіть папку, що містить аудіозаписи для перетворення.",
+ "Enter input path": "Введіть вхідний шлях",
+ "Output Folder": "Вихідна папка",
+ "Select the folder where the output audios will be saved.": "Виберіть папку, куди будуть збережені вихідні аудіозаписи.",
+ "Enter output path": "Введіть вихідний шлях",
+ "Get information about the audio": "Отримання інформації про аудіо",
+ "Information about the audio file": "Інформація про аудіофайл",
+ "Waiting for information...": "Чекаємо на інформацію...",
+ "## Voice Blender": "## Голосовий блендер",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Виберіть дві моделі голосу, встановіть бажаний відсоток накладання та змішайте їх у абсолютно новий голос.",
+ "Voice Blender": "Голосовий блендер",
+ "Drag and drop your model here": "Перетягніть модель сюди",
+ "You can also use a custom path.": "Ви також можете використовувати власний шлях.",
+ "Blend Ratio": "Пропорції змішування",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Регулювання положення більше в ту чи іншу сторону зробить модель більш схожою на першу або другу.",
+ "Fusion": "Злиття",
+ "Path to Model": "Шлях до моделі",
+ "Enter path to model": "Введіть шлях до моделі",
+ "Model information to be placed": "Інформація про модель, яку потрібно розмістити",
+ "Inroduce the model information": "Ознайомлення з інформацією про модель",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Інформація, яку потрібно розмістити в моделі (Ви можете залишити її порожньою або поставити що завгодно).",
+ "View model information": "Перегляд інформації про модель",
+ "Introduce the model pth path": "Ознайомлення з моделлю pth шляху",
+ "View": "Вид",
+ "Model extraction": "Вилучення моделі",
+ "Model conversion": "Перетворення моделі",
+ "Pth file": "Pth файл",
+ "Output of the pth file": "Висновок p-го файлу",
+ "# How to Report an Issue on GitHub": "# Як повідомити про проблему на GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Натисніть кнопку «Записати екран» нижче, щоб почати запис проблеми, з якою ви зіткнулися.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Після того, як ви закінчите запис випуску, натисніть кнопку «Зупинити запис» (та сама кнопка, але мітка змінюється залежно від того, активно ви записуєте чи ні).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Перейдіть до [GitHub Issues](https://github.com/IAHispano/Applio/issues) і натисніть кнопку «Новий випуск».",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Заповніть наданий шаблон проблеми, обов'язково включивши деталі за потреби, і скористайтеся розділом активів, щоб завантажити записаний файл з попереднього кроку.",
+ "Record Screen": "Екран запису",
+ "Record": "Запис",
+ "Stop Recording": "Зупинити записування",
+ "Introduce the model .pth path": "Представляємо модель .pth шляху",
+ "See Model Information": "Переглянути інформацію про модель",
+ "## Download Model": "## Завантажити модель",
+ "Model Link": "Посилання на модель",
+ "Introduce the model link": "Ознайомлення з моделлю",
+ "Download Model": "Завантажити модель",
+ "## Drop files": "## Скиньте файли",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Перетягніть файл .pth і файл .index у цей простір. Перетягніть одну, а потім іншу.",
+ "TTS Voices": "Голоси TTS",
+ "Select the TTS voice to use for the conversion.": "Виберіть голос TTS, який буде використано для перетворення.",
+ "Text to Synthesize": "Текст для синтезу",
+ "Enter the text to synthesize.": "Введіть текст для синтезу.",
+ "Or you can upload a .txt file": "Або ви можете завантажити файл .txt",
+ "Enter text to synthesize": "Введіть текст для синтезу",
+ "Output Path for TTS Audio": "Вихідний шлях для аудіо TTS",
+ "Output Path for RVC Audio": "Вихідний тракт для аудіо RVC",
+ "Enable Applio integration with Discord presence": "Увімкніть інтеграцію Applio з присутністю в Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Він активує можливість відображення поточної активності Applio в Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Увімкніть інтеграцію Applio з applio.org/models за допомогою колби",
+ "It will activate the possibility of downloading models with a click from the website.": "Він активує можливість завантаження моделей одним кліком з сайту.",
+ "Theme": "Тема",
+ "Select the theme you want to use. (Requires restarting Applio)": "Виберіть тему, яку потрібно використовувати. (Потрібно перезапустити Applio)",
+ "Language": "Мова",
+ "Select the language you want to use. (Requires restarting Applio)": "Виберіть потрібну мову. (Потрібно перезапустити Applio)",
+ "Plugin Installer": "Інсталятор плагінів",
+ "Drag your plugin.zip to install it": "Перетягніть plugin.zip, щоб встановити його",
+ "Version Checker": "Перевірка версій",
+ "Check which version of Applio is the latest to see if you need to update.": "Перевірте, яка версія Applio найновіша, щоб дізнатися, чи потрібно вам оновлюватися.",
+ "Check for updates": "Перевірте наявність оновлень"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/ur_UR.json b/assets/i18n/languages/ur_UR.json
new file mode 100644
index 0000000000000000000000000000000000000000..0df96313aaa5d2918c3d3eb0e22181c2dc6dc87a
--- /dev/null
+++ b/assets/i18n/languages/ur_UR.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "حتمی صوتی کلوننگ ٹول ، بے مثال طاقت ، ماڈیولریٹی ، اور صارف دوست تجربے کے لئے احتیاط سے بہتر بنایا گیا ہے۔",
+ "This section contains some extra utilities that often may be in experimental phases.": "اس سیکشن میں کچھ اضافی افادیت شامل ہیں جو اکثر تجرباتی مراحل میں ہوسکتی ہیں۔",
+ "Output Information": "آؤٹ پٹ معلومات",
+ "The output information will be displayed here.": "آؤٹ پٹ کی معلومات یہاں ظاہر کی جائے گی۔",
+ "Inference": "استدلال",
+ "Train": "ٹرین",
+ "Extra": "اضافی",
+ "Merge Audios": "آڈیو کو ضم کریں",
+ "Processing": "پروسیسنگ",
+ "Audio Analyzer": "Audio Analyzer",
+ "Model Information": "ماڈل کی معلومات",
+ "Plugins": "پلگ ان",
+ "Download": "ڈاؤن لوڈ",
+ "Report a Bug": "ایک بگ کی رپورٹ کریں",
+ "Settings": "سیٹنگیں",
+ "Preprocess": "پری پروسیس",
+ "Model Name": "ماڈل کا نام",
+ "Name of the new model.": "نئے ماڈل کا نام",
+ "Enter model name": "ماڈل کا نام درج کریں",
+ "Dataset Path": "ڈیٹا سیٹ کا راستہ",
+ "Path to the dataset folder.": "ڈیٹا سیٹ فولڈر کا راستہ۔",
+ "Refresh Datasets": "تازہ ترین ڈیٹا سیٹ",
+ "Dataset Creator": "ڈیٹا سیٹ تخلیق کار",
+ "Dataset Name": "ڈیٹا سیٹ کا نام",
+ "Name of the new dataset.": "نئے ڈیٹا سیٹ کا نام۔",
+ "Enter dataset name": "ڈیٹا سیٹ کا نام درج کریں",
+ "Upload Audio Dataset": "آڈیو ڈیٹا سیٹ اپ لوڈ کریں",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "آڈیو فائل کو کامیابی سے ڈیٹا سیٹ میں شامل کیا گیا ہے۔ براہ کرم پری پروسیس بٹن پر کلک کریں۔",
+ "Enter dataset path": "ڈیٹا سیٹ کا راستہ درج کریں",
+ "Sampling Rate": "نمونے لینے کی شرح",
+ "The sampling rate of the audio files.": "آڈیو فائلوں کے نمونے لینے کی شرح۔",
+ "RVC Version": "RVC Version",
+ "The RVC version of the model.": "ماڈل کا آر وی سی ورژن۔",
+ "Preprocess Dataset": "پری پروسیس ڈیٹا سیٹ",
+ "Extract": "نکالنا",
+ "Hop Length": "ہاپ کی لمبائی",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "نظام کو ایک اہم پچ تبدیلی میں منتقل ہونے میں لگنے والے عرصے کی نشاندہی کرتا ہے۔ چھوٹی ہاپ لمبائی کو اندازہ لگانے کے لئے زیادہ وقت کی ضرورت ہوتی ہے لیکن پچ کی درستگی زیادہ ہوتی ہے۔",
+ "Batch Size": "Batch کا سائز",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "اسے اپنے جی پی یو کے دستیاب وی آر اے ایم کے ساتھ ہم آہنگ کرنے کا مشورہ دیا جاتا ہے۔ 4 کی ترتیب بہتر درستگی لیکن سست پروسیسنگ پیش کرتی ہے ، جبکہ 8 تیز اور معیاری نتائج فراہم کرتی ہے۔",
+ "Save Every Epoch": "ہر دور کو محفوظ کریں",
+ "Determine at how many epochs the model will saved at.": "اس بات کا تعین کریں کہ ماڈل کتنے ادوار میں محفوظ رہے گا۔",
+ "Total Epoch": "مجموعی دور کی نوعیت",
+ "Specifies the overall quantity of epochs for the model training process.": "ماڈل ٹریننگ کے عمل کے لئے ادوار کی مجموعی مقدار کی وضاحت کرتا ہے۔",
+ "Pretrained": "پہلے سے تربیت یافتہ",
+ "Save Only Latest": "صرف تازہ ترین محفوظ کریں",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "اس ترتیب کو فعال کرنے کے نتیجے میں جی اور ڈی فائلیں صرف اپنے تازہ ترین ورژن کو محفوظ کریں گی ، مؤثر طریقے سے اسٹوریج کی جگہ کو محفوظ کریں گی۔",
+ "Save Every Weights": "ہر وزن کو بچائیں",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "یہ ترتیب آپ کو ہر دور کے اختتام پر ماڈل کے وزن کو بچانے کے قابل بناتی ہے۔",
+ "Custom Pretrained": "اپنی مرضی کے مطابق پہلے سے تربیت یافتہ",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "اپنی مرضی کے مطابق پہلے سے تربیت یافتہ ماڈلز کا استعمال بہتر نتائج کا باعث بن سکتا ہے ، کیونکہ مخصوص استعمال کے معاملے کے مطابق سب سے مناسب پری ٹریننگ ماڈلز کا انتخاب کارکردگی کو نمایاں طور پر بڑھا سکتا ہے۔",
+ "Upload Pretrained Model": "پہلے سے تربیت یافتہ ماڈل اپ لوڈ کریں",
+ "Refresh Custom Pretraineds": "اپنی مرضی کے مطابق پیشگی تربیت یافتہ تازہ کاری کریں",
+ "Pretrained Custom Settings": "پہلے سے تربیت یافتہ کسٹم ترتیبات",
+ "The file you dropped is not a valid pretrained file. Please try again.": "آپ نے جو فائل چھوڑی ہے وہ درست پہلے سے تربیت یافتہ فائل نہیں ہے۔ براہ مہربانی دوبارہ کوشش کریں۔",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "ڈراپ ڈاؤن مینو میں پہلے سے تربیت یافتہ فائل دیکھنے کے لئے ریفریش بٹن پر کلک کریں۔",
+ "Pretrained G Path": "اپنی مرضی کے مطابق پہلے سے تربیت یافتہ G",
+ "Pretrained D Path": "اپنی مرضی کے مطابق پہلے سے تربیت یافتہ D",
+ "GPU Settings": "GPU Settings",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "بہتر جی پی یو آرکیٹیکچر والے صارفین کے لئے تجویز کردہ جدید جی پی یو سیٹنگز سیٹ کرتا ہے۔",
+ "GPU Custom Settings": "GPU اپنی مرضی کے مطابق ترتیبات",
+ "GPU Number": "GPU نمبر",
+ "0 to ∞ separated by -": "0 سے الگ ∞ -",
+ "GPU Information": "GPU Information",
+ "Pitch Guidance": "پچ گائیڈنس",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "پچ گائیڈنس کا استعمال کرتے ہوئے ، اس کی پچ سمیت اصل آواز کے آئینے کو آئینہ دار بنانا ممکن ہوجاتا ہے۔ یہ خصوصیت خاص طور پر گانے اور دیگر منظرناموں کے لئے قابل قدر ہے جہاں اصل دھن یا پچ پیٹرن کو محفوظ کرنا ضروری ہے۔",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "اپنی تربیت کرتے وقت پہلے سے تربیت یافتہ ماڈلز کا استعمال کریں۔ یہ نقطہ نظر تربیت کے دورانیے کو کم کرتا ہے اور مجموعی معیار کو بڑھاتا ہے.",
+ "Extract Features": "نکالنے کی خصوصیات",
+ "Start Training": "تربیت شروع کریں",
+ "Generate Index": "انڈیکس پیدا کریں",
+ "Voice Model": "صوتی ماڈل",
+ "Select the voice model to use for the conversion.": "تبادلے کے لئے استعمال کرنے کے لئے صوتی ماڈل منتخب کریں۔",
+ "Index File": "انڈیکس فائل",
+ "Select the index file to use for the conversion.": "تبادلے کے لئے استعمال کرنے کے لئے انڈیکس فائل منتخب کریں۔",
+ "Refresh": "تازہ",
+ "Unload Voice": "آواز کو ان لوڈ کریں",
+ "Single": "تنہا",
+ "Upload Audio": "آڈیو اپ لوڈ کریں",
+ "Select Audio": "آڈیو منتخب کریں",
+ "Select the audio to convert.": "تبدیل کرنے کے لئے آڈیو منتخب کریں۔",
+ "Advanced Settings": "اعلی درجے کی ترتیبات",
+ "Clear Outputs (Deletes all audios in assets/audios)": "آؤٹ پٹ صاف کریں (اثاثوں / آڈیو میں تمام آڈیو حذف کرتا ہے)",
+ "Custom Output Path": "اپنی مرضی کے مطابق آؤٹ پٹ پتھ",
+ "Output Path": "آؤٹ پٹ پتھ",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "وہ راستہ جہاں آؤٹ پٹ آڈیو محفوظ کیا جائے گا ، اثاثوں / آڈیو / output.wav میں ڈیفالٹ طور پر",
+ "Split Audio": "آڈیو کو تقسیم کریں",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "کچھ معاملات میں بہتر نتائج حاصل کرنے کے لئے آڈیو کو ٹکڑوں میں تقسیم کریں۔",
+ "Autotune": "Autotune",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "گانے کی تبدیلیوں کے لئے تجویز کردہ اپنے تخمینوں پر نرم آٹو ٹیون لگائیں۔",
+ "Clean Audio": "صاف آڈیو",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "آواز کا پتہ لگانے والے الگورتھم کا استعمال کرتے ہوئے اپنے آڈیو آؤٹ پٹ کو صاف کریں ، جو آڈیو بولنے کے لئے سفارش کی جاتی ہے۔",
+ "Clean Strength": "صاف طاقت",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "آپ جس آڈیو کو چاہتے ہیں اس پر کلین اپ لیول سیٹ کریں ، جتنا زیادہ آپ اسے بڑھائیں گے اتنا ہی یہ صاف ہوجائے گا ، لیکن یہ ممکن ہے کہ آڈیو زیادہ کمپریسڈ ہوجائے۔",
+ "Pitch": "پچ",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "آڈیو کی پچ سیٹ کریں ، قیمت جتنی زیادہ ہوگی ، پچ اتنی ہی زیادہ ہوگی۔",
+ "Filter Radius": "فلٹر کے دائرے",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "اگر تعداد تین سے زیادہ یا اس کے برابر ہے تو ، جمع کردہ ٹون کے نتائج پر اوسط فلٹرنگ کا استعمال سانس کو کم کرنے کی صلاحیت رکھتا ہے۔",
+ "Search Feature Ratio": "تلاش کی خصوصیت کا تناسب",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "انڈیکس فائل کے ذریعہ اثر و رسوخ؛ ایک اعلی قیمت زیادہ اثر و رسوخ سے مطابقت رکھتی ہے. تاہم ، کم اقدار کا انتخاب آڈیو میں موجود نوادرات کو کم کرنے میں مدد کرسکتا ہے۔",
+ "Volume Envelope": "حجم کا لفافہ",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "آؤٹ پٹ کے حجم لفافے کے ساتھ متبادل یا مرکب کریں۔ تناسب 1 کے جتنا قریب ہوتا ہے ، اتنا ہی زیادہ آؤٹ پٹ لفافہ استعمال ہوتا ہے۔",
+ "Protect Voiceless Consonants": "آواز کے بغیر عبارتوں کی حفاظت کریں",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "الیکٹرو-صوتی پھٹنے اور دیگر نوادرات کو روکنے کے لئے مختلف عبارتوں اور سانس لینے کی آوازوں کی حفاظت کریں۔ پیرامیٹر کو اس کی زیادہ سے زیادہ قیمت 0.5 تک کھینچنا جامع تحفظ فراہم کرتا ہے۔ تاہم ، اس قدر کو کم کرنے سے تحفظ کی حد کم ہوسکتی ہے جبکہ ممکنہ طور پر انڈیکسنگ اثر کو کم کیا جاسکتا ہے۔",
+ "Pitch extraction algorithm": "پچ نکالنے کا الگورتھم",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "آڈیو تبدیلی کے لئے استعمال کرنے کے لئے پچ نکالنے کا الگورتھم۔ ڈیفالٹ الگورتھم آر ایم وی پی ای ہے ، جو زیادہ تر معاملات کے لئے سفارش کی جاتی ہے۔",
+ "Convert": "بدلیں",
+ "Export Audio": "آڈیو برآمد کریں",
+ "Batch": "بیچ",
+ "Input Folder": "ان پٹ فولڈر",
+ "Select the folder containing the audios to convert.": "تبدیل کرنے کے لئے آڈیو پر مشتمل فولڈر منتخب کریں۔",
+ "Enter input path": "ان پٹ راستہ درج کریں",
+ "Output Folder": "آؤٹ پٹ فولڈر",
+ "Select the folder where the output audios will be saved.": "وہ فولڈر منتخب کریں جہاں آؤٹ پٹ آڈیو محفوظ کیے جائیں گے۔",
+ "Enter output path": "آؤٹ پٹ کا راستہ درج کریں",
+ "Get information about the audio": "آڈیو کے بارے میں معلومات حاصل کریں",
+ "Information about the audio file": "آڈیو فائل کے بارے میں معلومات",
+ "Waiting for information...": "معلومات کا انتظار ہے...",
+ "## Voice Blender": "## وائس بلینڈر",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "دو صوتی ماڈل منتخب کریں ، اپنا مطلوبہ مرکب فیصد مقرر کریں ، اور انہیں مکمل طور پر نئی آواز میں بلینڈ کریں۔",
+ "Voice Blender": "Voice Blender",
+ "Drag and drop your model here": "اپنے ماڈل کو یہاں گھسیٹیں اور چھوڑیں",
+ "You can also use a custom path.": "آپ اپنی مرضی کے مطابق راستہ بھی استعمال کرسکتے ہیں۔",
+ "Blend Ratio": "مرکب تناسب",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "پوزیشن کو ایک طرف یا دوسری طرف زیادہ ایڈجسٹ کرنے سے ماڈل پہلے یا دوسرے سے زیادہ مماثلت رکھتا ہے۔",
+ "Fusion": "فیوژن",
+ "Path to Model": "ماڈل کا راستہ",
+ "Enter path to model": "ماڈل کا راستہ درج کریں",
+ "Model information to be placed": "ماڈل کی معلومات رکھی جائے گی",
+ "Inroduce the model information": "ماڈل کی معلومات فراہم کریں",
+ "The information to be placed in the model (You can leave it blank or put anything).": "ماڈل میں رکھی جانے والی معلومات (آپ اسے خالی چھوڑ سکتے ہیں یا کچھ بھی ڈال سکتے ہیں).",
+ "View model information": "ماڈل کی معلومات دیکھیں",
+ "Introduce the model pth path": "ماڈل پی ٹی ایچ راستہ متعارف کروائیں",
+ "View": "منظر",
+ "Model extraction": "ماڈل نکالنے",
+ "Model conversion": "ماڈل کی تبدیلی",
+ "Pth file": "پی ٹی ایچ فائل",
+ "Output of the pth file": "پی ٹی ایچ فائل کی آؤٹ پٹ",
+ "# How to Report an Issue on GitHub": "# گیٹ ہب پر کسی مسئلے کی اطلاع کیسے دیں",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. آپ جس مسئلے کا سامنا کر رہے ہیں اسے ریکارڈ کرنا شروع کرنے کے لئے نیچے 'ریکارڈ اسکرین' بٹن پر کلک کریں۔",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. ایک بار جب آپ مسئلے کی ریکارڈنگ مکمل کرلیں تو ، 'اسٹاپ ریکارڈنگ' بٹن پر کلک کریں (وہی بٹن ، لیکن لیبل اس بات پر منحصر ہے کہ آپ فعال طور پر ریکارڈنگ کر رہے ہیں یا نہیں)۔",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. [گیٹ ہب ایشوز] (https://github.com/IAHispano/Applio/issues) پر جائیں اور 'نیا مسئلہ' بٹن پر کلک کریں۔",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. فراہم کردہ ایشو ٹیمپلیٹ کو مکمل کریں ، ضرورت کے مطابق تفصیلات شامل کرنے کو یقینی بنائیں ، اور پچھلے مرحلے سے ریکارڈ شدہ فائل کو اپ لوڈ کرنے کے لئے اثاثوں کے سیکشن کا استعمال کریں۔",
+ "Record Screen": "ریکارڈ اسکرین",
+ "Record": "ریکارڈ",
+ "Stop Recording": "ریکارڈنگ بند کریں",
+ "Introduce the model .pth path": "ماڈل .pth پتھ متعارف کروائیں",
+ "See Model Information": "ماڈل کی معلومات دیکھیں",
+ "## Download Model": "## ڈاؤن لوڈ ماڈل",
+ "Model Link": "ماڈل لنک",
+ "Introduce the model link": "ماڈل کا لنک متعارف کروائیں",
+ "Download Model": "ڈاؤن لوڈ ماڈل",
+ "## Drop files": "## فائلیں چھوڑ دیں",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "اپنی .pth فائل اور .انڈیکس فائل کو اس جگہ میں گھسیٹیں۔ ایک کو گھسیٹیں اور پھر دوسرے کو۔",
+ "TTS Voices": "ٹی ٹی ایس وائسز",
+ "Select the TTS voice to use for the conversion.": "تبدیلی کے لئے استعمال کرنے کے لئے TTS آواز منتخب کریں۔",
+ "Text to Synthesize": "ترکیب کرنے کے لئے متن",
+ "Enter the text to synthesize.": "ترکیب کرنے کے لئے متن درج کریں۔",
+ "Or you can upload a .txt file": "یا آپ .txt فائل اپ لوڈ کرسکتے ہیں",
+ "Enter text to synthesize": "ترتیب دینے کے لئے متن درج کریں",
+ "Output Path for TTS Audio": "ٹی ٹی ایس آڈیو کے لئے آؤٹ پٹ پتھ",
+ "Output Path for RVC Audio": "آر وی سی آڈیو کے لئے آؤٹ پٹ پتھ",
+ "Enable Applio integration with Discord presence": "ڈسکارڈ کی موجودگی کے ساتھ ایپلیو انضمام کو فعال کریں",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "یہ ڈسکارڈ میں موجودہ ایپلیو سرگرمی کو ظاہر کرنے کے امکان کو فعال کرے گا۔",
+ "Enable Applio integration with applio.org/models using flask": "فلاسک کا استعمال کرتے ہوئے applio.org/models کے ساتھ ایپلیو انضمام کو فعال کریں",
+ "It will activate the possibility of downloading models with a click from the website.": "یہ ویب سائٹ سے ایک کلک کے ساتھ ماڈل ڈاؤن لوڈ کرنے کے امکان کو فعال کرے گا۔",
+ "Theme": "موضوع",
+ "Select the theme you want to use. (Requires restarting Applio)": "وہ تھیم منتخب کریں جسے آپ استعمال کرنا چاہتے ہیں۔ (ایپلیو کو دوبارہ شروع کرنے کی ضرورت ہے)",
+ "Language": "زبان",
+ "Select the language you want to use. (Requires restarting Applio)": "وہ زبان منتخب کریں جسے آپ استعمال کرنا چاہتے ہیں۔ (ایپلیو کو دوبارہ شروع کرنے کی ضرورت ہے)",
+ "Plugin Installer": "Plugin Installer",
+ "Drag your plugin.zip to install it": "اسے انسٹال کرنے کے لئے اپنے plugin.zip کو گھسیٹیں",
+ "Version Checker": "Version Checker",
+ "Check which version of Applio is the latest to see if you need to update.": "چیک کریں کہ ایپلیو کا کون سا ورژن تازہ ترین ہے یہ دیکھنے کے لئے کہ آیا آپ کو اپ ڈیٹ کرنے کی ضرورت ہے۔",
+ "Check for updates": "اپ ڈیٹس کے لئے چیک کریں"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/vi_VI.json b/assets/i18n/languages/vi_VI.json
new file mode 100644
index 0000000000000000000000000000000000000000..0fa26561bf6e9f81e5ae3b656d8ae4f0b77996ef
--- /dev/null
+++ b/assets/i18n/languages/vi_VI.json
@@ -0,0 +1,175 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "Công cụ nhân bản giọng nói tối ưu, được tối ưu hóa tỉ mỉ cho sức mạnh vô song, tính mô-đun và trải nghiệm thân thiện với người dùng.",
+ "This section contains some extra utilities that often may be in experimental phases.": "Phần này chứa một số tiện ích bổ sung thường có thể đang trong giai đoạn thử nghiệm.",
+ "Output Information": "Thông tin đầu ra",
+ "The output information will be displayed here.": "Thông tin đầu ra sẽ được hiển thị ở đây.",
+ "Inference": "Suy luận",
+ "Train": "Xe lửa",
+ "Extra": "Phụ",
+ "Merge Audios": "Hợp nhất âm thanh",
+ "Processing": "Xử lý",
+ "Audio Analyzer": "Máy phân tích âm thanh",
+ "Model Information": "Thông tin mô hình",
+ "Plugins": "Plugin",
+ "Download": "Tải xuống",
+ "Report a Bug": "Báo cáo lỗi",
+ "Settings": "Cài đặt",
+ "Preprocess": "Tiền xử lý",
+ "Model Name": "Tên Model",
+ "Name of the new model.": "Tên của mô hình mới.",
+ "Enter model name": "Nhập tên model",
+ "Dataset Path": "Đường dẫn tập dữ liệu",
+ "Path to the dataset folder.": "Đường dẫn đến thư mục tập dữ liệu.",
+ "Refresh Datasets": "Làm mới tập dữ liệu",
+ "Dataset Creator": "Trình tạo tập dữ liệu",
+ "Dataset Name": "Tên tập dữ liệu",
+ "Name of the new dataset.": "Tên của tập dữ liệu mới.",
+ "Enter dataset name": "Nhập tên tập dữ liệu",
+ "Upload Audio Dataset": "Tải lên tập dữ liệu âm thanh",
+ "The audio file has been successfully added to the dataset. Please click the preprocess button.": "Tệp âm thanh đã được thêm thành công vào tập dữ liệu. Vui lòng nhấp vào nút xử lý trước.",
+ "Enter dataset path": "Nhập đường dẫn tập dữ liệu",
+ "Sampling Rate": "Tỷ lệ lấy mẫu",
+ "The sampling rate of the audio files.": "Tốc độ lấy mẫu của các tệp âm thanh.",
+ "RVC Version": "Phiên bản RVC",
+ "The RVC version of the model.": "Phiên bản RVC của mô hình.",
+ "Preprocess Dataset": "Tập dữ liệu tiền xử lý",
+ "Extract": "Trích",
+ "Hop Length": "Chiều dài hop",
+ "Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy.": "Biểu thị khoảng thời gian cần thiết để hệ thống chuyển sang thay đổi cao độ đáng kể. Độ dài bước nhảy nhỏ hơn đòi hỏi nhiều thời gian hơn để suy luận nhưng có xu hướng mang lại độ chính xác cao độ cao hơn.",
+ "Batch Size": "Kích thước lô",
+ "It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results.": "Bạn nên căn chỉnh nó với VRAM có sẵn của GPU của bạn. Cài đặt 4 cung cấp độ chính xác được cải thiện nhưng xử lý chậm hơn, trong khi 8 cung cấp kết quả tiêu chuẩn và nhanh hơn.",
+ "Save Every Epoch": "Lưu mọi kỷ nguyên",
+ "Determine at how many epochs the model will saved at.": "Xác định mô hình sẽ lưu tại bao nhiêu kỷ nguyên.",
+ "Total Epoch": "Tổng kỷ nguyên",
+ "Specifies the overall quantity of epochs for the model training process.": "Chỉ định số lượng tổng thể của các kỷ nguyên cho quá trình đào tạo mô hình.",
+ "Pretrained": "Đào tạo trước",
+ "Save Only Latest": "Chỉ lưu mới nhất",
+ "Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space.": "Bật cài đặt này sẽ dẫn đến các tệp G và D chỉ lưu các phiên bản mới nhất của chúng, tiết kiệm hiệu quả dung lượng lưu trữ.",
+ "Save Every Weights": "Tiết kiệm mọi trọng lượng",
+ "This setting enables you to save the weights of the model at the conclusion of each epoch.": "Cài đặt này cho phép bạn lưu trọng số của mô hình khi kết thúc mỗi kỷ nguyên.",
+ "Custom Pretrained": "Đào tạo trước tùy chỉnh",
+ "Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance.": "Sử dụng các mô hình được đào tạo trước tùy chỉnh có thể dẫn đến kết quả vượt trội, vì việc lựa chọn các mô hình được đào tạo trước phù hợp nhất phù hợp với trường hợp sử dụng cụ thể có thể nâng cao đáng kể hiệu suất.",
+ "Upload Pretrained Model": "Tải lên mô hình được đào tạo trước",
+ "Refresh Custom Pretraineds": "Làm mới Custom Pretraineds",
+ "Pretrained Custom Settings": "Cài đặt tùy chỉnh được đào tạo sẵn",
+ "The file you dropped is not a valid pretrained file. Please try again.": "Tệp bạn đã bỏ không phải là tệp được đào tạo trước hợp lệ. Vui lòng thử lại.",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "Nhấp vào nút làm mới để xem tệp được đào tạo trước trong menu thả xuống.",
+ "Pretrained G Path": "Tùy chỉnh được đào tạo trước G",
+ "Pretrained D Path": "Tùy chỉnh được đào tạo trước D",
+ "GPU Settings": "Cài đặt GPU",
+ "Sets advanced GPU settings, recommended for users with better GPU architecture.": "Đặt cài đặt GPU nâng cao, được khuyến nghị cho người dùng có kiến trúc GPU tốt hơn.",
+ "GPU Custom Settings": "Cài đặt tùy chỉnh GPU",
+ "GPU Number": "Số GPU",
+ "0 to ∞ separated by -": "0 đến ∞ cách nhau bởi -",
+ "GPU Information": "Thông tin GPU",
+ "Pitch Guidance": "Hướng dẫn quảng cáo chiêu hàng",
+ "By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential.": "Bằng cách sử dụng hướng dẫn cao độ, nó trở nên khả thi để phản ánh ngữ điệu của giọng nói gốc, bao gồm cả cao độ của nó. Tính năng này đặc biệt có giá trị đối với ca hát và các tình huống khác trong đó việc giữ nguyên giai điệu hoặc cao độ ban đầu là điều cần thiết.",
+ "Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality.": "Sử dụng các mô hình được đào tạo trước khi đào tạo của riêng bạn. Cách tiếp cận này làm giảm thời gian đào tạo và nâng cao chất lượng tổng thể.",
+ "Extract Features": "Tính năng trích xuất",
+ "Start Training": "Bắt đầu đào tạo",
+ "Generate Index": "Tạo chỉ mục",
+ "Voice Model": "Mô hình giọng nói",
+ "Select the voice model to use for the conversion.": "Chọn kiểu giọng nói để sử dụng cho quá trình chuyển đổi.",
+ "Index File": "Tệp chỉ mục",
+ "Select the index file to use for the conversion.": "Chọn tệp chỉ mục để sử dụng cho quá trình chuyển đổi.",
+ "Refresh": "Làm tươi",
+ "Unload Voice": "Dỡ giọng nói",
+ "Single": "Đơn",
+ "Upload Audio": "Tải lên âm thanh",
+ "Select Audio": "Chọn Âm thanh",
+ "Select the audio to convert.": "Chọn âm thanh để chuyển đổi.",
+ "Advanced Settings": "Cài đặt nâng cao",
+ "Clear Outputs (Deletes all audios in assets/audios)": "Xóa đầu ra (Xóa tất cả âm thanh trong nội dung / âm thanh)",
+ "Custom Output Path": "Đường dẫn đầu ra tùy chỉnh",
+ "Output Path": "Đường dẫn đầu ra",
+ "The path where the output audio will be saved, by default in assets/audios/output.wav": "Đường dẫn nơi âm thanh đầu ra sẽ được lưu, theo mặc định trong tài sản / âm thanh / output.wav",
+ "Split Audio": "Tách âm thanh",
+ "Split the audio into chunks for inference to obtain better results in some cases.": "Chia âm thanh thành các phần để suy luận nhằm thu được kết quả tốt hơn trong một số trường hợp.",
+ "Autotune": "Tự động điều chỉnh",
+ "Apply a soft autotune to your inferences, recommended for singing conversions.": "Áp dụng autotune mềm cho suy luận của bạn, được đề xuất để chuyển đổi ca hát.",
+ "Clean Audio": "Âm thanh sạch",
+ "Clean your audio output using noise detection algorithms, recommended for speaking audios.": "Làm sạch đầu ra âm thanh của bạn bằng các thuật toán phát hiện tiếng ồn, được khuyến nghị để nói âm thanh.",
+ "Clean Strength": "Sức mạnh sạch",
+ "Set the clean-up level to the audio you want, the more you increase it the more it will clean up, but it is possible that the audio will be more compressed.": "Đặt mức dọn dẹp thành âm thanh bạn muốn, bạn càng tăng thì càng dọn dẹp, nhưng có thể âm thanh sẽ bị nén nhiều hơn.",
+ "Pitch": "Sân",
+ "Set the pitch of the audio, the higher the value, the higher the pitch.": "Đặt cao độ của âm thanh, giá trị càng cao, cao độ càng cao.",
+ "Filter Radius": "Bán kính lọc",
+ "If the number is greater than or equal to three, employing median filtering on the collected tone results has the potential to decrease respiration.": "Nếu số lượng lớn hơn hoặc bằng ba, việc sử dụng bộ lọc trung bình trên kết quả âm thu thập được có khả năng làm giảm hô hấp.",
+ "Search Feature Ratio": "Tỷ lệ tính năng tìm kiếm",
+ "Influence exerted by the index file; a higher value corresponds to greater influence. However, opting for lower values can help mitigate artifacts present in the audio.": "Ảnh hưởng tác động của tệp chỉ mục; Giá trị cao hơn tương ứng với ảnh hưởng lớn hơn. Tuy nhiên, việc chọn các giá trị thấp hơn có thể giúp giảm thiểu các hiện vật có trong âm thanh.",
+ "Volume Envelope": "Phong bì khối lượng",
+ "Substitute or blend with the volume envelope of the output. The closer the ratio is to 1, the more the output envelope is employed.": "Thay thế hoặc trộn với phong bì âm lượng của đầu ra. Tỷ lệ càng gần 1, phong bì đầu ra càng được sử dụng.",
+ "Protect Voiceless Consonants": "Bảo vệ phụ âm vô thanh",
+ "Safeguard distinct consonants and breathing sounds to prevent electro-acoustic tearing and other artifacts. Pulling the parameter to its maximum value of 0.5 offers comprehensive protection. However, reducing this value might decrease the extent of protection while potentially mitigating the indexing effect.": "Bảo vệ các phụ âm riêng biệt và âm thanh thở để ngăn ngừa rách âm thanh điện và các hiện vật khác. Kéo tham số đến giá trị tối đa 0, 5 cung cấp sự bảo vệ toàn diện. Tuy nhiên, việc giảm giá trị này có thể làm giảm mức độ bảo vệ trong khi có khả năng giảm thiểu hiệu ứng lập chỉ mục.",
+ "Pitch extraction algorithm": "Thuật toán trích xuất cao độ",
+ "Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases.": "Thuật toán trích xuất cao độ để sử dụng cho việc chuyển đổi âm thanh. Thuật toán mặc định là rmvpe, được khuyến nghị cho hầu hết các trường hợp.",
+ "Convert": "Convert",
+ "Export Audio": "Xuất âm thanh",
+ "Batch": "Mẻ",
+ "Input Folder": "Thư mục đầu vào",
+ "Select the folder containing the audios to convert.": "Chọn thư mục chứa âm thanh để chuyển đổi.",
+ "Enter input path": "Nhập đường dẫn nhập liệu",
+ "Output Folder": "Thư mục đầu ra",
+ "Select the folder where the output audios will be saved.": "Chọn thư mục lưu âm thanh đầu ra.",
+ "Enter output path": "Nhập đường dẫn đầu ra",
+ "Get information about the audio": "Nhận thông tin về âm thanh",
+ "Information about the audio file": "Thông tin về tệp âm thanh",
+ "Waiting for information...": "Đang chờ thông tin...",
+ "## Voice Blender": "## Máy xay sinh tố giọng nói",
+ "Select two voice models, set your desired blend percentage, and blend them into an entirely new voice.": "Chọn hai mẫu giọng nói, đặt tỷ lệ phần trăm pha trộn mong muốn của bạn và trộn chúng thành một giọng nói hoàn toàn mới.",
+ "Voice Blender": "Máy xay sinh tố giọng nói",
+ "Drag and drop your model here": "Kéo và thả mô hình của bạn vào đây",
+ "You can also use a custom path.": "Bạn cũng có thể sử dụng đường dẫn tùy chỉnh.",
+ "Blend Ratio": "Tỷ lệ pha trộn",
+ "Adjusting the position more towards one side or the other will make the model more similar to the first or second.": "Điều chỉnh vị trí nhiều hơn về phía bên này hay bên kia sẽ làm cho mô hình giống với thứ nhất hoặc thứ hai hơn.",
+ "Fusion": "Fusion",
+ "Path to Model": "Đường dẫn đến mô hình",
+ "Enter path to model": "Nhập đường dẫn đến mô hình",
+ "Model information to be placed": "Thông tin mô hình sẽ được đặt",
+ "Inroduce the model information": "Giới thiệu thông tin mô hình",
+ "The information to be placed in the model (You can leave it blank or put anything).": "Thông tin được đặt trong mô hình (Bạn có thể để trống hoặc đặt bất cứ thứ gì).",
+ "View model information": "Xem thông tin mô hình",
+ "Introduce the model pth path": "Giới thiệu mô hình pth path",
+ "View": "Cảnh",
+ "Model extraction": "Trích xuất mô hình",
+ "Model conversion": "Chuyển đổi mô hình",
+ "Pth file": "Tệp Pth",
+ "Output of the pth file": "Đầu ra của tệp pth",
+ "# How to Report an Issue on GitHub": "# Cách báo cáo sự cố trên GitHub",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1. Nhấp vào nút 'Ghi lại màn hình' bên dưới để bắt đầu ghi lại sự cố bạn đang gặp phải.",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. Khi bạn đã ghi xong sự cố, hãy nhấp vào nút 'Dừng ghi' (cùng một nút, nhưng nhãn thay đổi tùy thuộc vào việc bạn có chủ động ghi hay không).",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. Đi tới [Vấn đề GitHub] (https://github.com/IAHispano/Applio/issues) và nhấp vào nút 'Vấn đề mới'.",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. Hoàn thành mẫu vấn đề được cung cấp, đảm bảo bao gồm các chi tiết khi cần thiết và sử dụng phần tài sản để tải lên tệp đã ghi từ bước trước.",
+ "Record Screen": "Ghi lại màn hình",
+ "Record": "Ghi",
+ "Stop Recording": "Dừng ghi",
+ "Introduce the model .pth path": "Giới thiệu mô hình đường dẫn .pth",
+ "See Model Information": "Xem thông tin mô hình",
+ "## Download Model": "## Tải xuống mô hình",
+ "Model Link": "Liên kết mô hình",
+ "Introduce the model link": "Giới thiệu link mô hình",
+ "Download Model": "Tải xuống mô hình",
+ "## Drop files": "## Thả tệp",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "Kéo tệp .pth và tệp .index của bạn vào không gian này. Kéo cái này rồi cái kia.",
+ "TTS Voices": "Tiếng nói TTS",
+ "Select the TTS voice to use for the conversion.": "Chọn giọng nói TTS để sử dụng cho quá trình chuyển đổi.",
+ "Text to Synthesize": "Văn bản để tổng hợp",
+ "Enter the text to synthesize.": "Nhập văn bản để tổng hợp.",
+ "Or you can upload a .txt file": "Hoặc bạn có thể tải lên tệp .txt",
+ "Enter text to synthesize": "Nhập văn bản để tổng hợp",
+ "Output Path for TTS Audio": "Đường dẫn đầu ra cho âm thanh TTS",
+ "Output Path for RVC Audio": "Đường dẫn đầu ra cho âm thanh RVC",
+ "Enable Applio integration with Discord presence": "Bật tích hợp Applio với sự hiện diện của Discord",
+ "It will activate the possibility of displaying the current Applio activity in Discord.": "Nó sẽ kích hoạt khả năng hiển thị hoạt động Applio hiện tại trong Discord.",
+ "Enable Applio integration with applio.org/models using flask": "Bật tích hợp Applio với applio.org/models bằng bình",
+ "It will activate the possibility of downloading models with a click from the website.": "Nó sẽ kích hoạt khả năng tải xuống các mô hình bằng một cú nhấp chuột từ trang web.",
+ "Theme": "Đề tài",
+ "Select the theme you want to use. (Requires restarting Applio)": "Chọn chủ đề bạn muốn sử dụng. (Yêu cầu khởi động lại Applio)",
+ "Language": "Ngôn ngữ",
+ "Select the language you want to use. (Requires restarting Applio)": "Chọn ngôn ngữ bạn muốn sử dụng. (Yêu cầu khởi động lại Applio)",
+ "Plugin Installer": "Trình cài đặt plugin",
+ "Drag your plugin.zip to install it": "Kéo plugin.zip của bạn để cài đặt nó",
+ "Version Checker": "Trình kiểm tra phiên bản",
+ "Check which version of Applio is the latest to see if you need to update.": "Kiểm tra xem phiên bản Applio nào là phiên bản mới nhất để xem bạn có cần cập nhật hay không.",
+ "Check for updates": "Kiểm tra bản cập nhật"
+}
\ No newline at end of file
diff --git a/assets/i18n/languages/wu_WU.json b/assets/i18n/languages/wu_WU.json
new file mode 100644
index 0000000000000000000000000000000000000000..9e26dfeeb6e641a33dae4961196235bdb965b21b
--- /dev/null
+++ b/assets/i18n/languages/wu_WU.json
@@ -0,0 +1 @@
+{}
\ No newline at end of file
diff --git a/assets/i18n/languages/zh_CN.json b/assets/i18n/languages/zh_CN.json
new file mode 100644
index 0000000000000000000000000000000000000000..78feed4dcb2fbf5028561549dc1d191b35c35288
--- /dev/null
+++ b/assets/i18n/languages/zh_CN.json
@@ -0,0 +1,113 @@
+{
+ "Ultimate voice cloning tool, meticulously optimized for unrivaled power, modularity, and user-friendly experience.": "终极语音克隆工具，经过精心优化，具有无与伦比的功能、模块化和用户友好的体验。",
+ "This section contains some extra utilities that often may be in experimental phases.": "本节包含一些额外的实用程序，这些实用程序通常处于实验阶段。",
+ "Output Information": "输出信息",
+ "Inference": "推理",
+ "Train": "火车",
+ "Extra": "额外",
+ "Merge Audios": "合并音频",
+ "Processing": "加工",
+ "Audio Analyzer": "音频分析仪",
+ "Model Information": "型号信息",
+ "Download": "下载",
+ "Report a Bug": "报告错误",
+ "Preprocess": "预处理",
+ "Model Name": "型号名称",
+ "Enter model name": "输入型号名称",
+ "Dataset Path": "数据集路径",
+ "Enter dataset path": "输入数据集路径",
+ "Sampling Rate": "采样率",
+ "RVC Version": "RVC 版本",
+ "Preprocess Dataset": "预处理数据集",
+ "Extract": "提取",
+ "Hop Length": "跳跃长度",
+ "Batch Size": "批量大小",
+ "Save Every Epoch": "保存每个纪元",
+ "Total Epoch": "总纪元",
+ "Pretrained": "预训练",
+ "Save Only Latest": "仅保存最新",
+ "Save Every Weights": "节省每一次砝码",
+ "Custom Pretrained": "自定义预训练",
+ "Upload Pretrained Model": "上传预训练模型",
+ "Pretrained Custom Settings": "预训练的自定义设置",
+ "The file you dropped is not a valid pretrained file. Please try again.": "您删除的文件不是有效的预训练文件。请再试一次。",
+ "Click the refresh button to see the pretrained file in the dropdown menu.": "单击刷新按钮，在下拉菜单中查看预训练文件。",
+ "Pretrained G Path": "自定义预训练 G",
+ "Pretrained D Path": "自定义预训练 D",
+ "GPU Settings": "GPU 设置",
+ "GPU Custom Settings": "GPU 自定义设置",
+ "GPU Number": "GPU 数量",
+ "0 to ∞ separated by -": "0 到 ∞ 之间用 -",
+ "GPU Information": "GPU 信息",
+ "Pitch Guidance": "音高指导",
+ "Extract Features": "提取特征",
+ "Start Training": "开始训练",
+ "Generate Index": "生成索引",
+ "Voice Model": "语音模型",
+ "Index File": "Index 文件",
+ "Refresh": "刷新",
+ "Unload Voice": "卸载语音",
+ "Single": "单",
+ "Upload Audio": "上传音频",
+ "Select Audio": "选择音频",
+ "Advanced Settings": "高级设置",
+ "Clear Outputs (Deletes all audios in assets/audios)": "清除输出（删除资产/音频中的所有音频）",
+ "Custom Output Path": "自定义输出路径",
+ "Output Path": "输出路径",
+ "Pitch": "投",
+ "If >=3: apply median filtering to the harvested pitch results. The value represents the filter radius and can reduce breathiness": "如果 >=3：对收获的音高结果应用中值滤波。该值表示过滤器半径，可以减少呼吸",
+ "Search Feature Ratio": "搜索特征比率",
+ "Pitch extraction algorithm": "音高提取算法",
+ "Convert": "转换",
+ "Export Audio": "导出音频",
+ "Batch": "批",
+ "Input Folder": "输入文件夹",
+ "Enter input path": "输入输入路径",
+ "Output Folder": "输出文件夹",
+ "Enter output path": "输入输出路径",
+ "Get information about the audio": "获取有关音频的信息",
+ "Information about the audio file": "有关音频文件的信息",
+ "Waiting for information...": "等待信息...",
+ "Model fusion": "模型融合",
+ "Weight for Model A": "A型重量",
+ "Whether the model has pitch guidance": "模型是否具有俯仰引导",
+ "Model architecture version": "模型架构版本",
+ "Path to Model A": "模型 A 的路径",
+ "Path to Model B": "模型 B 的路径",
+ "Path to model": "模型路径",
+ "Model information to be placed": "要放置的模型信息",
+ "Fusion": "融合",
+ "Modify model information": "修改模型信息",
+ "Path to Model": "模型路径",
+ "Model information to be modified": "要修改的模型信息",
+ "Save file name": "保存文件名",
+ "Modify": "修改",
+ "View model information": "查看型号信息",
+ "View": "视图",
+ "Model extraction": "模型提取",
+ "Model conversion": "模型转换",
+ "Pth file": "Pth 文件",
+ "Output of the pth file": "pth 文件的输出",
+ "# How to Report an Issue on GitHub": "# 如何在 GitHub 上报告问题",
+ "1. Click on the 'Record Screen' button below to start recording the issue you are experiencing.": "1.单击下面的“录制屏幕”按钮开始记录您遇到的问题。",
+ "2. Once you have finished recording the issue, click on the 'Stop Recording' button (the same button, but the label changes depending on whether you are actively recording or not).": "2. 录制完问题后，单击“停止录制”按钮（相同的按钮，但标签会根据您是否正在录制而变化）。",
+ "3. Go to [GitHub Issues](https://github.com/IAHispano/Applio/issues) and click on the 'New Issue' button.": "3. 转到 [GitHub 问题]（https://github.com/IAHispano/Applio/issues），然后单击“新问题”按钮。",
+ "4. Complete the provided issue template, ensuring to include details as needed, and utilize the assets section to upload the recorded file from the previous step.": "4. 填写提供的问题模板，确保根据需要包含详细信息，并利用资产部分上传上一步的记录文件。",
+ "Record Screen": "录制屏幕",
+ "Record": "记录",
+ "Stop Recording": "停止录制",
+ "Introduce the model .pth path": "引入模型 .pth 路径",
+ "See Model Information": "查看型号信息",
+ "## Download Model": "## 下载模型",
+ "Model Link": "模型链接",
+ "Introduce the model link": "介绍模型链接",
+ "Download Model": "下载模型",
+ "## Drop files": "## 删除文件",
+ "Drag your .pth file and .index file into this space. Drag one and then the other.": "将 .pth 文件和 .index 文件拖到此空间中。拖动一个，然后拖动另一个。",
+ "TTS Voices": "TTS语音",
+ "Text to Synthesize": "要合成的文本",
+ "Enter text to synthesize": "输入要合成的文本",
+ "Output Path for TTS Audio": "TTS 音频的输出路径",
+ "Output Path for RVC Audio": "RVC 音频的输出路径",
+ "Enable Applio integration with Discord presence": "Applio 存在"
+}
diff --git a/i18n/scan_i18n.py b/assets/i18n/scan.py
similarity index 50%
rename from i18n/scan_i18n.py
rename to assets/i18n/scan.py
index f3e52cf4f9f06d78877d77d2353f666aa759e36f..7cd584fa2af8480f443c51417c442fecf5197d11 100644
--- a/i18n/scan_i18n.py
+++ b/assets/i18n/scan.py
@@ -1,75 +1,71 @@
-import ast
-import glob
-import json
-from collections import OrderedDict
-
-
-def extract_i18n_strings(node):
- i18n_strings = []
-
- if (
- isinstance(node, ast.Call)
- and isinstance(node.func, ast.Name)
- and node.func.id == "i18n"
- ):
- for arg in node.args:
- if isinstance(arg, ast.Str):
- i18n_strings.append(arg.s)
-
- for child_node in ast.iter_child_nodes(node):
- i18n_strings.extend(extract_i18n_strings(child_node))
-
- return i18n_strings
-
-
-# scan the directory for all .py files (recursively)
-# for each file, parse the code into an AST
-# for each AST, extract the i18n strings
-
-strings = []
-for filename in glob.iglob("**/*.py", recursive=True):
- with open(filename, "r") as f:
- code = f.read()
- if "I18nAuto" in code:
- tree = ast.parse(code)
- i18n_strings = extract_i18n_strings(tree)
- print(filename, len(i18n_strings))
- strings.extend(i18n_strings)
-code_keys = set(strings)
-"""
-n_i18n.py
-gui_v1.py 26
-app.py 16
-infer-web.py 147
-scan_i18n.py 0
-i18n.py 0
-lib/train/process_ckpt.py 1
-"""
-print()
-print("Total unique:", len(code_keys))
-
-
-standard_file = "i18n/locale/zh_CN.json"
-with open(standard_file, "r", encoding="utf-8") as f:
- standard_data = json.load(f, object_pairs_hook=OrderedDict)
-standard_keys = set(standard_data.keys())
-
-# Define the standard file name
-unused_keys = standard_keys - code_keys
-print("Unused keys:", len(unused_keys))
-for unused_key in unused_keys:
- print("\t", unused_key)
-
-missing_keys = code_keys - standard_keys
-print("Missing keys:", len(missing_keys))
-for missing_key in missing_keys:
- print("\t", missing_key)
-
-code_keys_dict = OrderedDict()
-for s in strings:
- code_keys_dict[s] = s
-
-# write back
-with open(standard_file, "w", encoding="utf-8") as f:
- json.dump(code_keys_dict, f, ensure_ascii=False, indent=4, sort_keys=True)
- f.write("\n")
+import ast
+import json
+from pathlib import Path
+from collections import OrderedDict
+
+
+def extract_i18n_strings(node):
+ i18n_strings = []
+
+ if (
+ isinstance(node, ast.Call)
+ and isinstance(node.func, ast.Name)
+ and node.func.id == "i18n"
+ ):
+ for arg in node.args:
+ if isinstance(arg, ast.Str):
+ i18n_strings.append(arg.s)
+
+ for child_node in ast.iter_child_nodes(node):
+ i18n_strings.extend(extract_i18n_strings(child_node))
+
+ return i18n_strings
+
+
+def process_file(file_path):
+ with open(file_path, "r", encoding="utf8") as file:
+ code = file.read()
+ if "I18nAuto" in code:
+ tree = ast.parse(code)
+ i18n_strings = extract_i18n_strings(tree)
+ print(file_path, len(i18n_strings))
+ return i18n_strings
+ return []
+
+
+# Use pathlib for file handling
+py_files = Path(".").rglob("*.py")
+
+# Use a set to store unique strings
+code_keys = set()
+
+for py_file in py_files:
+ strings = process_file(py_file)
+ code_keys.update(strings)
+
+print()
+print("Total unique:", len(code_keys))
+
+standard_file = "languages/en_US.json"
+with open(standard_file, "r", encoding="utf-8") as file:
+ standard_data = json.load(file, object_pairs_hook=OrderedDict)
+standard_keys = set(standard_data.keys())
+
+# Combine unused and missing keys sections
+unused_keys = standard_keys - code_keys
+missing_keys = code_keys - standard_keys
+
+print("Unused keys:", len(unused_keys))
+for unused_key in unused_keys:
+ print("\t", unused_key)
+
+print("Missing keys:", len(missing_keys))
+for missing_key in missing_keys:
+ print("\t", missing_key)
+
+code_keys_dict = OrderedDict((s, s) for s in code_keys)
+
+# Use context manager for writing back to the file
+with open(standard_file, "w", encoding="utf-8") as file:
+ json.dump(code_keys_dict, file, ensure_ascii=False, indent=4, sort_keys=True)
+ file.write("\n")
diff --git a/assets/installation_checker.py b/assets/installation_checker.py
new file mode 100644
index 0000000000000000000000000000000000000000..44cbfe72d620adb7add6a41dd7ce0499ef2af071
--- /dev/null
+++ b/assets/installation_checker.py
@@ -0,0 +1,38 @@
+import sys
+import os
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+
+
+class InstallationError(Exception):
+ def __init__(self, message="InstallationError"):
+ self.message = message
+ super().__init__(self.message)
+
+
+def check_installation():
+ try:
+ system_drive = os.getenv("SystemDrive")
+ current_drive = os.path.splitdrive(now_dir)[0]
+ if current_drive.upper() != system_drive.upper():
+ raise InstallationError(
+ f"Error: Current working directory is not on the default system drive ({system_drive}). Please move Applio in the correct drive."
+ )
+ except:
+ pass
+ else:
+ if "OneDrive" in now_dir:
+ raise InstallationError(
+ "Error: Current working directory is on OneDrive. Please move Applio in another folder."
+ )
+ elif " " in now_dir:
+ raise InstallationError(
+ "Error: Current working directory contains spaces. Please move Applio in another folder."
+ )
+ try:
+ now_dir.encode("ascii")
+ except UnicodeEncodeError:
+ raise InstallationError(
+ "Error: Current working directory contains non-ASCII characters. Please move Applio in another folder."
+ )
diff --git a/assets/pretrained/.gitignore b/assets/pretrained/.gitignore
deleted file mode 100644
index d6b7ef32c8478a48c3994dcadc86837f4371184d..0000000000000000000000000000000000000000
--- a/assets/pretrained/.gitignore
+++ /dev/null
@@ -1,2 +0,0 @@
-*
-!.gitignore
diff --git a/assets/pretrained_v2/.gitignore b/assets/pretrained_v2/.gitignore
deleted file mode 100644
index d6b7ef32c8478a48c3994dcadc86837f4371184d..0000000000000000000000000000000000000000
--- a/assets/pretrained_v2/.gitignore
+++ /dev/null
@@ -1,2 +0,0 @@
-*
-!.gitignore
diff --git a/assets/rmvpe/.gitignore b/assets/rmvpe/.gitignore
deleted file mode 100644
index d6b7ef32c8478a48c3994dcadc86837f4371184d..0000000000000000000000000000000000000000
--- a/assets/rmvpe/.gitignore
+++ /dev/null
@@ -1,2 +0,0 @@
-*
-!.gitignore
diff --git a/assets/themes/Applio.py b/assets/themes/Applio.py
new file mode 100644
index 0000000000000000000000000000000000000000..b171b3f639bc283be9f2ff150f60a9a50ff0d2e9
--- /dev/null
+++ b/assets/themes/Applio.py
@@ -0,0 +1,284 @@
+from __future__ import annotations
+
+from typing import Iterable
+import gradio as gr
+
+# gr.themes.builder()
+from gradio.themes.base import Base
+from gradio.themes.utils import colors, fonts, sizes
+import time
+
+
+class Applio(Base):
+ def __init__(
+ self,
+ *,
+ primary_hue: colors.Color | str = colors.neutral,
+ secondary_hue: colors.Color | str = colors.neutral,
+ neutral_hue: colors.Color | str = colors.neutral,
+ spacing_size: sizes.Size | str = sizes.spacing_md,
+ radius_size: sizes.Size | str = sizes.radius_md,
+ text_size: sizes.Size | str = sizes.text_lg,
+ font: fonts.Font | str | Iterable[fonts.Font | str] = (
+ "Syne V",
+ fonts.GoogleFont("Syne"),
+ "ui-sans-serif",
+ "system-ui",
+ ),
+ font_mono: fonts.Font | str | Iterable[fonts.Font | str] = (
+ "ui-monospace",
+ fonts.GoogleFont("Nunito Sans"),
+ ),
+ ):
+ super().__init__(
+ primary_hue=primary_hue,
+ secondary_hue=secondary_hue,
+ neutral_hue=neutral_hue,
+ spacing_size=spacing_size,
+ radius_size=radius_size,
+ text_size=text_size,
+ font=font,
+ font_mono=font_mono,
+ )
+ self.name = ("Applio",)
+ self.secondary_100 = ("#dbeafe",)
+ self.secondary_200 = ("#bfdbfe",)
+ self.secondary_300 = ("#93c5fd",)
+ self.secondary_400 = ("#60a5fa",)
+ self.secondary_50 = ("#eff6ff",)
+ self.secondary_500 = ("#3b82f6",)
+ self.secondary_600 = ("#2563eb",)
+ self.secondary_700 = ("#1d4ed8",)
+ self.secondary_800 = ("#1e40af",)
+ self.secondary_900 = ("#1e3a8a",)
+ self.secondary_950 = ("#1d3660",)
+
+ super().set(
+ # Blaise
+ background_fill_primary="#110F0F",
+ background_fill_primary_dark="#110F0F",
+ background_fill_secondary="#110F0F",
+ background_fill_secondary_dark="#110F0F",
+ block_background_fill="*neutral_800",
+ block_background_fill_dark="*neutral_800",
+ block_border_color="*border_color_primary",
+ block_border_color_dark="*border_color_primary",
+ block_border_width="1px",
+ block_border_width_dark="1px",
+ block_info_text_color="*body_text_color_subdued",
+ block_info_text_color_dark="*body_text_color_subdued",
+ block_info_text_size="*text_sm",
+ block_info_text_weight="400",
+ block_label_background_fill="*background_fill_primary",
+ block_label_background_fill_dark="*background_fill_secondary",
+ block_label_border_color="*border_color_primary",
+ block_label_border_color_dark="*border_color_primary",
+ block_label_border_width="1px",
+ block_label_border_width_dark="1px",
+ block_label_margin="0",
+ block_label_padding="*spacing_sm *spacing_lg",
+ block_label_radius="calc(*radius_lg - 1px) 0 calc(*radius_lg - 1px) 0",
+ block_label_right_radius="0 calc(*radius_lg - 1px) 0 calc(*radius_lg - 1px)",
+ block_label_shadow="*block_shadow",
+ block_label_text_color="*#110F0F",
+ block_label_text_color_dark="*#110F0F",
+ block_label_text_weight="400",
+ block_padding="*spacing_xl",
+ block_radius="*radius_md",
+ block_shadow="none",
+ block_shadow_dark="none",
+ block_title_background_fill="rgb(255,255,255)",
+ block_title_background_fill_dark="rgb(255,255,255)",
+ block_title_border_color="none",
+ block_title_border_color_dark="none",
+ block_title_border_width="0px",
+ block_title_padding="*block_label_padding",
+ block_title_radius="*block_label_radius",
+ block_title_text_color="#110F0F",
+ block_title_text_color_dark="#110F0F",
+ block_title_text_size="*text_md",
+ block_title_text_weight="600",
+ body_background_fill="#110F0F",
+ body_background_fill_dark="#110F0F",
+ body_text_color="white",
+ body_text_color_dark="white",
+ body_text_color_subdued="*neutral_400",
+ body_text_color_subdued_dark="*neutral_400",
+ body_text_size="*text_md",
+ body_text_weight="400",
+ border_color_accent="*neutral_600",
+ border_color_accent_dark="*neutral_600",
+ border_color_primary="*neutral_800",
+ border_color_primary_dark="*neutral_800",
+ button_border_width="*input_border_width",
+ button_border_width_dark="*input_border_width",
+ button_cancel_background_fill="*button_secondary_background_fill",
+ button_cancel_background_fill_dark="*button_secondary_background_fill",
+ button_cancel_background_fill_hover="*button_cancel_background_fill",
+ button_cancel_background_fill_hover_dark="*button_cancel_background_fill",
+ button_cancel_border_color="*button_secondary_border_color",
+ button_cancel_border_color_dark="*button_secondary_border_color",
+ button_cancel_border_color_hover="*button_cancel_border_color",
+ button_cancel_border_color_hover_dark="*button_cancel_border_color",
+ button_cancel_text_color="#110F0F",
+ button_cancel_text_color_dark="#110F0F",
+ button_cancel_text_color_hover="#110F0F",
+ button_cancel_text_color_hover_dark="#110F0F",
+ button_large_padding="*spacing_lg calc(2 * *spacing_lg)",
+ button_large_radius="*radius_lg",
+ button_large_text_size="*text_lg",
+ button_large_text_weight="600",
+ button_primary_background_fill="*primary_600",
+ button_primary_background_fill_dark="*primary_600",
+ button_primary_background_fill_hover="*primary_500",
+ button_primary_background_fill_hover_dark="*primary_500",
+ button_primary_border_color="*primary_500",
+ button_primary_border_color_dark="*primary_500",
+ button_primary_border_color_hover="*primary_400",
+ button_primary_border_color_hover_dark="*primary_400",
+ button_primary_text_color="white",
+ button_primary_text_color_dark="white",
+ button_primary_text_color_hover="#110F0F",
+ button_primary_text_color_hover_dark="#110F0F",
+ button_secondary_background_fill="transparent",
+ button_secondary_background_fill_dark="transparent",
+ button_secondary_background_fill_hover="*neutral_800",
+ button_secondary_background_fill_hover_dark="*neutral_800",
+ button_secondary_border_color="*neutral_700",
+ button_secondary_border_color_dark="*neutral_700",
+ button_secondary_border_color_hover="*neutral_600",
+ button_secondary_border_color_hover_dark="*neutral_600",
+ button_secondary_text_color="white",
+ button_secondary_text_color_dark="white",
+ button_secondary_text_color_hover="*button_secondary_text_color",
+ button_secondary_text_color_hover_dark="*button_secondary_text_color",
+ button_shadow="none",
+ button_shadow_active="*shadow_inset",
+ button_shadow_hover="none",
+ button_small_padding="*spacing_sm calc(2 * *spacing_sm)",
+ button_small_radius="*radius_lg",
+ button_small_text_size="*text_md",
+ button_small_text_weight="400",
+ button_transition="0.3s ease all",
+ checkbox_background_color="*neutral_700",
+ checkbox_background_color_dark="*neutral_700",
+ checkbox_background_color_focus="*checkbox_background_color",
+ checkbox_background_color_focus_dark="*checkbox_background_color",
+ checkbox_background_color_hover="*checkbox_background_color",
+ checkbox_background_color_hover_dark="*checkbox_background_color",
+ checkbox_background_color_selected="*secondary_600",
+ checkbox_background_color_selected_dark="*secondary_600",
+ checkbox_border_color="*neutral_700",
+ checkbox_border_color_dark="*neutral_700",
+ checkbox_border_color_focus="*secondary_500",
+ checkbox_border_color_focus_dark="*secondary_500",
+ checkbox_border_color_hover="*neutral_600",
+ checkbox_border_color_hover_dark="*neutral_600",
+ checkbox_border_color_selected="*secondary_600",
+ checkbox_border_color_selected_dark="*secondary_600",
+ checkbox_border_radius="*radius_sm",
+ checkbox_border_width="*input_border_width",
+ checkbox_border_width_dark="*input_border_width",
+ checkbox_check="url(\"data:image/svg+xml,%3csvg viewBox='0 0 16 16' fill='white' xmlns='http://www.w3.org/2000/svg'%3e%3cpath d='M12.207 4.793a1 1 0 010 1.414l-5 5a1 1 0 01-1.414 0l-2-2a1 1 0 011.414-1.414L6.5 9.086l4.293-4.293a1 1 0 011.414 0z'/%3e%3c/svg%3e\")",
+ checkbox_label_background_fill="transparent",
+ checkbox_label_background_fill_dark="transparent",
+ checkbox_label_background_fill_hover="transparent",
+ checkbox_label_background_fill_hover_dark="transparent",
+ checkbox_label_background_fill_selected="transparent",
+ checkbox_label_background_fill_selected_dark="transparent",
+ checkbox_label_border_color="transparent",
+ checkbox_label_border_color_dark="transparent",
+ checkbox_label_border_color_hover="transparent",
+ checkbox_label_border_color_hover_dark="transparent",
+ checkbox_label_border_width="transparent",
+ checkbox_label_border_width_dark="transparent",
+ checkbox_label_gap="*spacing_lg",
+ checkbox_label_padding="*spacing_md calc(2 * *spacing_md)",
+ checkbox_label_shadow="none",
+ checkbox_label_text_color="*body_text_color",
+ checkbox_label_text_color_dark="*body_text_color",
+ checkbox_label_text_color_selected="*checkbox_label_text_color",
+ checkbox_label_text_color_selected_dark="*checkbox_label_text_color",
+ checkbox_label_text_size="*text_md",
+ checkbox_label_text_weight="400",
+ checkbox_shadow="*input_shadow",
+ color_accent="*primary_500",
+ color_accent_soft="*primary_50",
+ color_accent_soft_dark="*neutral_700",
+ container_radius="*radius_xl",
+ embed_radius="*radius_lg",
+ error_background_fill="*background_fill_primary",
+ error_background_fill_dark="*background_fill_primary",
+ error_border_color="*border_color_primary",
+ error_border_color_dark="*border_color_primary",
+ error_border_width="1px",
+ error_border_width_dark="1px",
+ error_text_color="#ef4444",
+ error_text_color_dark="#ef4444",
+ form_gap_width="0px",
+ input_background_fill="*neutral_900",
+ input_background_fill_dark="*neutral_900",
+ input_background_fill_focus="*secondary_600",
+ input_background_fill_focus_dark="*secondary_600",
+ input_background_fill_hover="*input_background_fill",
+ input_background_fill_hover_dark="*input_background_fill",
+ input_border_color="*neutral_700",
+ input_border_color_dark="*neutral_700",
+ input_border_color_focus="*secondary_600",
+ input_border_color_focus_dark="*primary_600",
+ input_border_color_hover="*input_border_color",
+ input_border_color_hover_dark="*input_border_color",
+ input_border_width="1px",
+ input_border_width_dark="1px",
+ input_padding="*spacing_xl",
+ input_placeholder_color="*neutral_500",
+ input_placeholder_color_dark="*neutral_500",
+ input_radius="*radius_lg",
+ input_shadow="none",
+ input_shadow_dark="none",
+ input_shadow_focus="*input_shadow",
+ input_shadow_focus_dark="*input_shadow",
+ input_text_size="*text_md",
+ input_text_weight="400",
+ layout_gap="*spacing_xxl",
+ link_text_color="*secondary_500",
+ link_text_color_active="*secondary_500",
+ link_text_color_active_dark="*secondary_500",
+ link_text_color_dark="*secondary_500",
+ link_text_color_hover="*secondary_400",
+ link_text_color_hover_dark="*secondary_400",
+ link_text_color_visited="*secondary_600",
+ link_text_color_visited_dark="*secondary_600",
+ loader_color="*color_accent",
+ loader_color_dark="*color_accent",
+ panel_background_fill="*background_fill_secondary",
+ panel_background_fill_dark="*background_fill_secondary",
+ panel_border_color="*border_color_primary",
+ panel_border_color_dark="*border_color_primary",
+ panel_border_width="1px",
+ panel_border_width_dark="1px",
+ prose_header_text_weight="600",
+ prose_text_size="*text_md",
+ prose_text_weight="400",
+ radio_circle="url(\"data:image/svg+xml,%3csvg viewBox='0 0 16 16' fill='white' xmlns='http://www.w3.org/2000/svg'%3e%3ccircle cx='8' cy='8' r='3'/%3e%3c/svg%3e\")",
+ section_header_text_size="*text_md",
+ section_header_text_weight="400",
+ shadow_drop="rgba(0,0,0,0.05) 0px 1px 2px 0px",
+ shadow_drop_lg="0 1px 3px 0 rgb(0 0 0 / 0.1), 0 1px 2px -1px rgb(0 0 0 / 0.1)",
+ shadow_inset="rgba(0,0,0,0.05) 0px 2px 4px 0px inset",
+ shadow_spread="3px",
+ shadow_spread_dark="1px",
+ slider_color="#9E9E9E",
+ slider_color_dark="#9E9E9E",
+ stat_background_fill="*primary_500",
+ stat_background_fill_dark="*primary_500",
+ table_border_color="*neutral_700",
+ table_border_color_dark="*neutral_700",
+ table_even_background_fill="*neutral_950",
+ table_even_background_fill_dark="*neutral_950",
+ table_odd_background_fill="*neutral_900",
+ table_odd_background_fill_dark="*neutral_900",
+ table_radius="*radius_lg",
+ table_row_focus="*color_accent_soft",
+ table_row_focus_dark="*color_accent_soft",
+ )
diff --git a/assets/themes/loadThemes.py b/assets/themes/loadThemes.py
new file mode 100644
index 0000000000000000000000000000000000000000..d95b4a39dddfa8eec7b694574f741269439d47f8
--- /dev/null
+++ b/assets/themes/loadThemes.py
@@ -0,0 +1,122 @@
+import json
+import os
+import importlib
+import gradio as gr
+
+now_dir = os.getcwd()
+
+folder = os.path.dirname(os.path.abspath(__file__))
+folder = os.path.dirname(folder)
+folder = os.path.dirname(folder)
+folder = os.path.join(folder, "assets", "themes")
+config_file = os.path.join(now_dir, "assets", "config.json")
+
+import sys
+
+sys.path.append(folder)
+
+
+def get_class(filename):
+ with open(filename, "r", encoding="utf8") as file:
+ for line_number, line in enumerate(file, start=1):
+ if "class " in line:
+ found = line.split("class ")[1].split(":")[0].split("(")[0].strip()
+ return found
+ break
+ return None
+
+
+def get_list():
+
+ themes_from_files = [
+ os.path.splitext(name)[0]
+ for root, _, files in os.walk(folder, topdown=False)
+ for name in files
+ if name.endswith(".py") and root == folder
+ ]
+
+ json_file_path = os.path.join(folder, "theme_list.json")
+
+ try:
+ with open(json_file_path, "r", encoding="utf8") as json_file:
+ themes_from_url = [item["id"] for item in json.load(json_file)]
+ except FileNotFoundError:
+ themes_from_url = []
+
+ combined_themes = set(themes_from_files + themes_from_url)
+
+ return list(combined_themes)
+
+
+def select_theme(name):
+ selected_file = name + ".py"
+ full_path = os.path.join(folder, selected_file)
+
+ if not os.path.exists(full_path):
+ with open(config_file, "r", encoding="utf8") as json_file:
+ config_data = json.load(json_file)
+
+ config_data["theme"]["file"] = None
+ config_data["theme"]["class"] = name
+
+ with open(config_file, "w", encoding="utf8") as json_file:
+ json.dump(config_data, json_file, indent=2)
+ print(f"Theme {name} successfully selected, restart applio.")
+ gr.Info(f"Theme {name} successfully selected, restart applio.")
+ return
+
+ class_found = get_class(full_path)
+ if class_found:
+ with open(config_file, "r", encoding="utf8") as json_file:
+ config_data = json.load(json_file)
+
+ config_data["theme"]["file"] = selected_file
+ config_data["theme"]["class"] = class_found
+
+ with open(config_file, "w", encoding="utf8") as json_file:
+ json.dump(config_data, json_file, indent=2)
+ print(f"Theme {name} successfully selected, restart applio.")
+ gr.Info(f"Theme {name} successfully selected, restart applio.")
+ else:
+ print(f"Theme {name} was not found.")
+
+
+def read_json():
+ try:
+ with open(config_file, "r", encoding="utf8") as json_file:
+ data = json.load(json_file)
+ selected_file = data["theme"]["file"]
+ class_name = data["theme"]["class"]
+
+ if selected_file is not None and class_name:
+ return class_name
+ elif selected_file == None and class_name:
+ return class_name
+ else:
+ return "ParityError/Interstellar"
+ except Exception as e:
+ print(f"Error reading config.json: {e}")
+ return "ParityError/Interstellar"
+
+
+def load_json():
+ try:
+ with open(config_file, "r", encoding="utf8") as json_file:
+ data = json.load(json_file)
+ selected_file = data["theme"]["file"]
+ class_name = data["theme"]["class"]
+
+ if selected_file is not None and class_name:
+ module = importlib.import_module(selected_file[:-3])
+ obtained_class = getattr(module, class_name)
+ instance = obtained_class()
+ print(f"Theme Loaded: {class_name}")
+ return instance
+ elif selected_file == None and class_name:
+ return class_name
+ else:
+ print("The theme is incorrect.")
+ return None
+ except Exception as e:
+ print(f"Error Loading: {str(e)}")
+ return None
diff --git a/assets/themes/theme.json b/assets/themes/theme.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8c42ff064e327b1364eb90d6f73bc4283a7c1bd
--- /dev/null
+++ b/assets/themes/theme.json
@@ -0,0 +1 @@
+{"file": "Applio.py", "class": "Applio"}
\ No newline at end of file
diff --git a/assets/themes/theme_list.json b/assets/themes/theme_list.json
new file mode 100644
index 0000000000000000000000000000000000000000..9de2456d2c2151e7d6765a9b5e0664e09f2041f7
--- /dev/null
+++ b/assets/themes/theme_list.json
@@ -0,0 +1,81 @@
+[
+ {"id": "freddyaboulton/dracula_revamped"},
+ {"id": "freddyaboulton/bad-theme-space"},
+ {"id": "gradio/dracula_revamped"},
+ {"id": "abidlabs/dracula_revamped"},
+ {"id": "gradio/dracula_test"},
+ {"id": "abidlabs/dracula_test"},
+ {"id": "gradio/seafoam"},
+ {"id": "gradio/glass"},
+ {"id": "gradio/monochrome"},
+ {"id": "gradio/soft"},
+ {"id": "gradio/default"},
+ {"id": "gradio/base"},
+ {"id": "abidlabs/pakistan"},
+ {"id": "dawood/microsoft_windows"},
+ {"id": "ysharma/steampunk"},
+ {"id": "ysharma/huggingface"},
+ {"id": "gstaff/xkcd"},
+ {"id": "JohnSmith9982/small_and_pretty"},
+ {"id": "abidlabs/Lime"},
+ {"id": "freddyaboulton/this-theme-does-not-exist-2"},
+ {"id": "aliabid94/new-theme"},
+ {"id": "aliabid94/test2"},
+ {"id": "aliabid94/test3"},
+ {"id": "aliabid94/test4"},
+ {"id": "abidlabs/banana"},
+ {"id": "freddyaboulton/test-blue"},
+ {"id": "gstaff/sketch"},
+ {"id": "gstaff/whiteboard"},
+ {"id": "ysharma/llamas"},
+ {"id": "abidlabs/font-test"},
+ {"id": "YenLai/Superhuman"},
+ {"id": "bethecloud/storj_theme"},
+ {"id": "sudeepshouche/minimalist"},
+ {"id": "knotdgaf/gradiotest"},
+ {"id": "ParityError/Interstellar"},
+ {"id": "ParityError/Anime"},
+ {"id": "Ajaxon6255/Emerald_Isle"},
+ {"id": "ParityError/LimeFace"},
+ {"id": "finlaymacklon/smooth_slate"},
+ {"id": "finlaymacklon/boxy_violet"},
+ {"id": "derekzen/stardust"},
+ {"id": "EveryPizza/Cartoony-Gradio-Theme"},
+ {"id": "Ifeanyi/Cyanister"},
+ {"id": "Tshackelton/IBMPlex-DenseReadable"},
+ {"id": "snehilsanyal/scikit-learn"},
+ {"id": "Himhimhim/xkcd"},
+ {"id": "shivi/calm_seafoam"},
+ {"id": "nota-ai/theme"},
+ {"id": "rawrsor1/Everforest"},
+ {"id": "SebastianBravo/simci_css"},
+ {"id": "rottenlittlecreature/Moon_Goblin"},
+ {"id": "abidlabs/test-yellow"},
+ {"id": "abidlabs/test-yellow3"},
+ {"id": "idspicQstitho/dracula_revamped"},
+ {"id": "kfahn/AnimalPose"},
+ {"id": "HaleyCH/HaleyCH_Theme"},
+ {"id": "simulKitke/dracula_test"},
+ {"id": "braintacles/CrimsonNight"},
+ {"id": "wentaohe/whiteboardv2"},
+ {"id": "reilnuud/polite"},
+ {"id": "remilia/Ghostly"},
+ {"id": "Franklisi/darkmode"},
+ {"id": "coding-alt/soft"},
+ {"id": "xiaobaiyuan/theme_land"},
+ {"id": "step-3-profit/Midnight-Deep"},
+ {"id": "xiaobaiyuan/theme_demo"},
+ {"id": "Taithrah/Minimal"},
+ {"id": "Insuz/SimpleIndigo"},
+ {"id": "zkunn/Alipay_Gradio_theme"},
+ {"id": "Insuz/Mocha"},
+ {"id": "xiaobaiyuan/theme_brief"},
+ {"id": "Ama434/434-base-Barlow"},
+ {"id": "Ama434/def_barlow"},
+ {"id": "Ama434/neutral-barlow"},
+ {"id": "dawood/dracula_test"},
+ {"id": "nuttea/Softblue"},
+ {"id": "BlueDancer/Alien_Diffusion"},
+ {"id": "naughtondale/monochrome"},
+ {"id": "Dagfinn1962/standard"}
+]
\ No newline at end of file
diff --git a/assets/uvr5_weights/.gitignore b/assets/uvr5_weights/.gitignore
deleted file mode 100644
index d6b7ef32c8478a48c3994dcadc86837f4371184d..0000000000000000000000000000000000000000
--- a/assets/uvr5_weights/.gitignore
+++ /dev/null
@@ -1,2 +0,0 @@
-*
-!.gitignore
diff --git a/assets/version_checker.py b/assets/version_checker.py
new file mode 100644
index 0000000000000000000000000000000000000000..356bf32370c6eaaac7aeef3f22dacb1986468114
--- /dev/null
+++ b/assets/version_checker.py
@@ -0,0 +1,44 @@
+import os, sys
+import json
+import requests
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+
+config_file = os.path.join(now_dir, "assets", "config.json")
+
+
+def load_local_version():
+ with open(config_file, "r", encoding="utf8") as file:
+ config = json.load(file)
+ return config["version"]
+
+
+def obtain_tag_name():
+ url = "https://api.github.com/repos/IAHispano/Applio/releases/latest"
+
+ try:
+ response = requests.get(url)
+ response.raise_for_status()
+
+ data = response.json()
+ tag_name = data["tag_name"]
+
+ return tag_name
+
+ except requests.exceptions.RequestException as e:
+ print(f"Error: {e}")
+ return None
+
+
+def compare_version():
+ local_version = load_local_version()
+ online_version = obtain_tag_name()
+ elements_online_version = list(map(int, online_version.split(".")))
+ elements_local_version = list(map(int, local_version.split(".")))
+
+ for online, local in zip(elements_online_version, elements_local_version):
+ if local < online:
+ return f"Your local {local_version} version is older than {online_version} the latest version"
+
+ return f"Your local version {local_version} is the latest version."
diff --git a/assets/weights/.gitignore b/assets/weights/.gitignore
deleted file mode 100644
index d6b7ef32c8478a48c3994dcadc86837f4371184d..0000000000000000000000000000000000000000
--- a/assets/weights/.gitignore
+++ /dev/null
@@ -1,2 +0,0 @@
-*
-!.gitignore
diff --git a/audioEffects.py b/audioEffects.py
deleted file mode 100644
index 1830b19e1a5e3ec1f431388d8444ef3a2c9ed91f..0000000000000000000000000000000000000000
--- a/audioEffects.py
+++ /dev/null
@@ -1,37 +0,0 @@
-from pedalboard import Pedalboard, Compressor, Reverb, NoiseGate
-from pedalboard.io import AudioFile
-import sys
-import os
-now_dir = os.getcwd()
-sys.path.append(now_dir)
-from i18n import I18nAuto
-i18n = I18nAuto()
-from pydub import AudioSegment
-import numpy as np
-import soundfile as sf
-from pydub.playback import play
-
-def process_audio(input_path, output_path, reverb_enabled, compressor_enabled, noise_gate_enabled, ):
- print(reverb_enabled)
- print(compressor_enabled)
- print(noise_gate_enabled)
- effects = []
- if reverb_enabled:
- effects.append(Reverb(room_size=0.01))
- if compressor_enabled:
- effects.append(Compressor(threshold_db=-10, ratio=25))
- if noise_gate_enabled:
- effects.append(NoiseGate(threshold_db=-16, ratio=1.5, release_ms=250))
-
- board = Pedalboard(effects)
-
- with AudioFile(input_path) as f:
- with AudioFile(output_path, 'w', f.samplerate, f.num_channels) as o:
- while f.tell() < f.frames:
- chunk = f.read(f.samplerate)
- effected = board(chunk, f.samplerate, reset=False)
- o.write(effected)
-
- result = i18n("Processed audio saved at: ") + output_path
- print(result)
- return output_path
\ No newline at end of file
diff --git a/colab_for_mdx.py b/colab_for_mdx.py
deleted file mode 100644
index 274846d0b5395865a05fce0da86b96d26ac06999..0000000000000000000000000000000000000000
--- a/colab_for_mdx.py
+++ /dev/null
@@ -1,71 +0,0 @@
-import json
-import os
-import gc
-import psutil
-import requests
-import subprocess
-import time
-import logging
-import sys
-import shutil
-now_dir = os.getcwd()
-sys.path.append(now_dir)
-first_cell_executed = False
-file_folder = "Colab-for-MDX_B"
-def first_cell_ran():
- global first_cell_executed
- if first_cell_executed:
- #print("The 'first_cell_ran' function has already been executed.")
- return
-
-
-
- first_cell_executed = True
- os.makedirs("tmp_models", exist_ok=True)
-
-
-
- class hide_opt: # hide outputs
- def __enter__(self):
- self._original_stdout = sys.stdout
- sys.stdout = open(os.devnull, "w")
-
- def __exit__(self, exc_type, exc_val, exc_tb):
- sys.stdout.close()
- sys.stdout = self._original_stdout
-
- def get_size(bytes, suffix="B"): # read ram
- global svmem
- factor = 1024
- for unit in ["", "K", "M", "G", "T", "P"]:
- if bytes < factor:
- return f"{bytes:.2f}{unit}{suffix}"
- bytes /= factor
- svmem = psutil.virtual_memory()
-
-
- def use_uvr_without_saving():
- print("Notice: files won't be saved to personal drive.")
- print(f"Downloading {file_folder}...", end=" ")
- with hide_opt():
- #os.chdir(mounting_path)
- items_to_move = ["demucs", "diffq","julius","model","separated","tracks","mdx.py","MDX-Net_Colab.ipynb"]
- subprocess.run(["git", "clone", "https://github.com/NaJeongMo/Colab-for-MDX_B.git"])
- for item_name in items_to_move:
- item_path = os.path.join(file_folder, item_name)
- if os.path.exists(item_path):
- if os.path.isfile(item_path):
- shutil.move(item_path, now_dir)
- elif os.path.isdir(item_path):
- shutil.move(item_path, now_dir)
- try:
- shutil.rmtree(file_folder)
- except PermissionError:
- print(f"No se pudo eliminar la carpeta {file_folder}. Puede estar relacionada con Git.")
-
-
- use_uvr_without_saving()
- print("done!")
- if not os.path.exists("tracks"):
- os.mkdir("tracks")
-first_cell_ran()
\ No newline at end of file
diff --git a/configs/32k.json b/configs/32k.json
deleted file mode 100644
index bcae72223ec09dc199009d7cb5ed405a0c0981cf..0000000000000000000000000000000000000000
--- a/configs/32k.json
+++ /dev/null
@@ -1,50 +0,0 @@
-{
- "train": {
- "log_interval": 200,
- "seed": 1234,
- "epochs": 20000,
- "learning_rate": 1e-4,
- "betas": [0.8, 0.99],
- "eps": 1e-9,
- "batch_size": 4,
- "fp16_run": false,
- "lr_decay": 0.999875,
- "segment_size": 12800,
- "init_lr_ratio": 1,
- "warmup_epochs": 0,
- "c_mel": 45,
- "c_kl": 1.0
- },
- "data": {
- "max_wav_value": 32768.0,
- "sampling_rate": 32000,
- "filter_length": 1024,
- "hop_length": 320,
- "win_length": 1024,
- "n_mel_channels": 80,
- "mel_fmin": 0.0,
- "mel_fmax": null
- },
- "model": {
- "inter_channels": 192,
- "hidden_channels": 192,
- "filter_channels": 768,
- "n_heads": 2,
- "n_layers": 6,
- "kernel_size": 3,
- "p_dropout": 0,
- "resblock": "1",
- "resblock_kernel_sizes": [3, 7, 11],
- "resblock_dilation_sizes": [
- [1, 3, 5],
- [1, 3, 5],
- [1, 3, 5]
- ],
- "upsample_rates": [10, 4, 2, 2, 2],
- "upsample_initial_channel": 512,
- "upsample_kernel_sizes": [16, 16, 4, 4, 4],
- "use_spectral_norm": false,
- "gin_channels": 256,
- "spk_embed_dim": 109
- }
-}
diff --git a/configs/32k_v2.json b/configs/32k_v2.json
deleted file mode 100644
index ad42f87b15e1ea68eff0a90db50fbc08d56c7aa9..0000000000000000000000000000000000000000
--- a/configs/32k_v2.json
+++ /dev/null
@@ -1,50 +0,0 @@
-{
- "train": {
- "log_interval": 200,
- "seed": 1234,
- "epochs": 20000,
- "learning_rate": 1e-4,
- "betas": [0.8, 0.99],
- "eps": 1e-9,
- "batch_size": 4,
- "fp16_run": true,
- "lr_decay": 0.999875,
- "segment_size": 12800,
- "init_lr_ratio": 1,
- "warmup_epochs": 0,
- "c_mel": 45,
- "c_kl": 1.0
- },
- "data": {
- "max_wav_value": 32768.0,
- "sampling_rate": 32000,
- "filter_length": 1024,
- "hop_length": 320,
- "win_length": 1024,
- "n_mel_channels": 80,
- "mel_fmin": 0.0,
- "mel_fmax": null
- },
- "model": {
- "inter_channels": 192,
- "hidden_channels": 192,
- "filter_channels": 768,
- "n_heads": 2,
- "n_layers": 6,
- "kernel_size": 3,
- "p_dropout": 0,
- "resblock": "1",
- "resblock_kernel_sizes": [3, 7, 11],
- "resblock_dilation_sizes": [
- [1, 3, 5],
- [1, 3, 5],
- [1, 3, 5]
- ],
- "upsample_rates": [10, 8, 2, 2],
- "upsample_initial_channel": 512,
- "upsample_kernel_sizes": [20, 16, 4, 4],
- "use_spectral_norm": false,
- "gin_channels": 256,
- "spk_embed_dim": 109
- }
-}
diff --git a/configs/40k.json b/configs/40k.json
deleted file mode 100644
index 28ff4d91f2618497fb39ad27872151bcb0d51761..0000000000000000000000000000000000000000
--- a/configs/40k.json
+++ /dev/null
@@ -1,50 +0,0 @@
-{
- "train": {
- "log_interval": 200,
- "seed": 1234,
- "epochs": 20000,
- "learning_rate": 1e-4,
- "betas": [0.8, 0.99],
- "eps": 1e-9,
- "batch_size": 4,
- "fp16_run": false,
- "lr_decay": 0.999875,
- "segment_size": 12800,
- "init_lr_ratio": 1,
- "warmup_epochs": 0,
- "c_mel": 45,
- "c_kl": 1.0
- },
- "data": {
- "max_wav_value": 32768.0,
- "sampling_rate": 40000,
- "filter_length": 2048,
- "hop_length": 400,
- "win_length": 2048,
- "n_mel_channels": 125,
- "mel_fmin": 0.0,
- "mel_fmax": null
- },
- "model": {
- "inter_channels": 192,
- "hidden_channels": 192,
- "filter_channels": 768,
- "n_heads": 2,
- "n_layers": 6,
- "kernel_size": 3,
- "p_dropout": 0,
- "resblock": "1",
- "resblock_kernel_sizes": [3, 7, 11],
- "resblock_dilation_sizes": [
- [1, 3, 5],
- [1, 3, 5],
- [1, 3, 5]
- ],
- "upsample_rates": [10, 10, 2, 2],
- "upsample_initial_channel": 512,
- "upsample_kernel_sizes": [16, 16, 4, 4],
- "use_spectral_norm": false,
- "gin_channels": 256,
- "spk_embed_dim": 109
- }
-}
diff --git a/configs/48k.json b/configs/48k.json
deleted file mode 100644
index 4d01946ed50ade92c1f85b548ab008a4cd617eb8..0000000000000000000000000000000000000000
--- a/configs/48k.json
+++ /dev/null
@@ -1,50 +0,0 @@
-{
- "train": {
- "log_interval": 200,
- "seed": 1234,
- "epochs": 20000,
- "learning_rate": 1e-4,
- "betas": [0.8, 0.99],
- "eps": 1e-9,
- "batch_size": 4,
- "fp16_run": false,
- "lr_decay": 0.999875,
- "segment_size": 11520,
- "init_lr_ratio": 1,
- "warmup_epochs": 0,
- "c_mel": 45,
- "c_kl": 1.0
- },
- "data": {
- "max_wav_value": 32768.0,
- "sampling_rate": 48000,
- "filter_length": 2048,
- "hop_length": 480,
- "win_length": 2048,
- "n_mel_channels": 128,
- "mel_fmin": 0.0,
- "mel_fmax": null
- },
- "model": {
- "inter_channels": 192,
- "hidden_channels": 192,
- "filter_channels": 768,
- "n_heads": 2,
- "n_layers": 6,
- "kernel_size": 3,
- "p_dropout": 0,
- "resblock": "1",
- "resblock_kernel_sizes": [3, 7, 11],
- "resblock_dilation_sizes": [
- [1, 3, 5],
- [1, 3, 5],
- [1, 3, 5]
- ],
- "upsample_rates": [10, 6, 2, 2, 2],
- "upsample_initial_channel": 512,
- "upsample_kernel_sizes": [16, 16, 4, 4, 4],
- "use_spectral_norm": false,
- "gin_channels": 256,
- "spk_embed_dim": 109
- }
-}
diff --git a/configs/48k_v2.json b/configs/48k_v2.json
deleted file mode 100644
index 50f06421912e2cd1f69768c35272981d75d86983..0000000000000000000000000000000000000000
--- a/configs/48k_v2.json
+++ /dev/null
@@ -1,50 +0,0 @@
-{
- "train": {
- "log_interval": 200,
- "seed": 1234,
- "epochs": 20000,
- "learning_rate": 1e-4,
- "betas": [0.8, 0.99],
- "eps": 1e-9,
- "batch_size": 4,
- "fp16_run": true,
- "lr_decay": 0.999875,
- "segment_size": 17280,
- "init_lr_ratio": 1,
- "warmup_epochs": 0,
- "c_mel": 45,
- "c_kl": 1.0
- },
- "data": {
- "max_wav_value": 32768.0,
- "sampling_rate": 48000,
- "filter_length": 2048,
- "hop_length": 480,
- "win_length": 2048,
- "n_mel_channels": 128,
- "mel_fmin": 0.0,
- "mel_fmax": null
- },
- "model": {
- "inter_channels": 192,
- "hidden_channels": 192,
- "filter_channels": 768,
- "n_heads": 2,
- "n_layers": 6,
- "kernel_size": 3,
- "p_dropout": 0,
- "resblock": "1",
- "resblock_kernel_sizes": [3, 7, 11],
- "resblock_dilation_sizes": [
- [1, 3, 5],
- [1, 3, 5],
- [1, 3, 5]
- ],
- "upsample_rates": [12, 10, 2, 2],
- "upsample_initial_channel": 512,
- "upsample_kernel_sizes": [24, 20, 4, 4],
- "use_spectral_norm": false,
- "gin_channels": 256,
- "spk_embed_dim": 109
- }
-}
diff --git a/configs/config.json b/configs/config.json
deleted file mode 100644
index 8e9c17669bf8028653fbfa5c9eeac23ec76c1cc9..0000000000000000000000000000000000000000
--- a/configs/config.json
+++ /dev/null
@@ -1,15 +0,0 @@
-{
- "pth_path": "assets/weights/kikiV1.pth",
- "index_path": "logs/kikiV1.index",
- "sg_input_device": "VoiceMeeter Output (VB-Audio Vo (MME)",
- "sg_output_device": "VoiceMeeter Aux Input (VB-Audio (MME)",
- "threhold": -45.0,
- "pitch": 12.0,
- "index_rate": 0.0,
- "rms_mix_rate": 0.0,
- "block_time": 0.25,
- "crossfade_length": 0.04,
- "extra_time": 2.0,
- "n_cpu": 6.0,
- "f0method": "rmvpe"
-}
diff --git a/configs/config.py b/configs/config.py
deleted file mode 100644
index e3b0205a1f0d62f674b9c3de2c5ab7ee90464945..0000000000000000000000000000000000000000
--- a/configs/config.py
+++ /dev/null
@@ -1,265 +0,0 @@
-import argparse
-import os
-import sys
-import json
-from multiprocessing import cpu_count
-
-import torch
-
-try:
- import intel_extension_for_pytorch as ipex # pylint: disable=import-error, unused-import
- if torch.xpu.is_available():
- from infer.modules.ipex import ipex_init
- ipex_init()
-except Exception:
- pass
-
-import logging
-
-logger = logging.getLogger(__name__)
-
-
-version_config_list = [
- "v1/32k.json",
- "v1/40k.json",
- "v1/48k.json",
- "v2/48k.json",
- "v2/32k.json",
-]
-
-
-def singleton_variable(func):
- def wrapper(*args, **kwargs):
- if not wrapper.instance:
- wrapper.instance = func(*args, **kwargs)
- return wrapper.instance
-
- wrapper.instance = None
- return wrapper
-
-
-@singleton_variable
-class Config:
- def __init__(self):
- self.device = "cuda:0"
- self.is_half = True
- self.n_cpu = 0
- self.gpu_name = None
- self.json_config = self.load_config_json()
- self.gpu_mem = None
- (
- self.python_cmd,
- self.listen_port,
- self.iscolab,
- self.noparallel,
- self.noautoopen,
- self.paperspace,
- self.is_cli,
- self.grtheme,
- self.dml,
- ) = self.arg_parse()
- self.instead = ""
- self.x_pad, self.x_query, self.x_center, self.x_max = self.device_config()
-
- @staticmethod
- def load_config_json() -> dict:
- d = {}
- for config_file in version_config_list:
- with open(f"configs/{config_file}", "r") as f:
- d[config_file] = json.load(f)
- return d
-
- @staticmethod
- def arg_parse() -> tuple:
- exe = sys.executable or "python"
- parser = argparse.ArgumentParser()
- parser.add_argument("--port", type=int, default=7865, help="Listen port")
- parser.add_argument("--pycmd", type=str, default=exe, help="Python command")
- parser.add_argument("--colab", action="store_true", help="Launch in colab")
- parser.add_argument(
- "--noparallel", action="store_true", help="Disable parallel processing"
- )
- parser.add_argument(
- "--noautoopen",
- action="store_true",
- help="Do not open in browser automatically",
- )
- parser.add_argument(
- "--paperspace",
- action="store_true",
- help="Note that this argument just shares a gradio link for the web UI. Thus can be used on other non-local CLI systems.",
- )
- parser.add_argument(
- "--is_cli",
- action="store_true",
- help="Use the CLI instead of setting up a gradio UI. This flag will launch an RVC text interface where you can execute functions from infer-web.py!",
- )
-
- parser.add_argument(
- "-t",
- "--theme",
- help = "Theme for Gradio. Format - `JohnSmith9982/small_and_pretty` (no backticks)",
- default = "JohnSmith9982/small_and_pretty",
- type = str
- )
-
- parser.add_argument(
- "--dml",
- action="store_true",
- help="Use DirectML backend instead of CUDA."
- )
-
- cmd_opts = parser.parse_args()
-
- cmd_opts.port = cmd_opts.port if 0 <= cmd_opts.port <= 65535 else 7865
-
- return (
- cmd_opts.pycmd,
- cmd_opts.port,
- cmd_opts.colab,
- cmd_opts.noparallel,
- cmd_opts.noautoopen,
- cmd_opts.paperspace,
- cmd_opts.is_cli,
- cmd_opts.theme,
- cmd_opts.dml,
- )
-
- # has_mps is only available in nightly pytorch (for now) and MasOS 12.3+.
- # check `getattr` and try it for compatibility
- @staticmethod
- def has_mps() -> bool:
- if not torch.backends.mps.is_available():
- return False
- try:
- torch.zeros(1).to(torch.device("mps"))
- return True
- except Exception:
- return False
-
- @staticmethod
- def has_xpu() -> bool:
- if hasattr(torch, "xpu") and torch.xpu.is_available():
- return True
- else:
- return False
-
- def use_fp32_config(self):
- for config_file in version_config_list:
- self.json_config[config_file]["train"]["fp16_run"] = False
-
- def device_config(self) -> tuple:
- if torch.cuda.is_available():
- if self.has_xpu():
- self.device = self.instead = "xpu:0"
- self.is_half = True
- i_device = int(self.device.split(":")[-1])
- self.gpu_name = torch.cuda.get_device_name(i_device)
- if (
- ("16" in self.gpu_name and "V100" not in self.gpu_name.upper())
- or "P40" in self.gpu_name.upper()
- or "P10" in self.gpu_name.upper()
- or "1060" in self.gpu_name
- or "1070" in self.gpu_name
- or "1080" in self.gpu_name
- ):
- logger.info("Found GPU %s, force to fp32", self.gpu_name)
- self.is_half = False
- self.use_fp32_config()
- else:
- logger.info("Found GPU %s", self.gpu_name)
- self.gpu_mem = int(
- torch.cuda.get_device_properties(i_device).total_memory
- / 1024
- / 1024
- / 1024
- + 0.4
- )
- if self.gpu_mem <= 4:
- with open("infer/modules/train/preprocess.py", "r") as f:
- strr = f.read().replace("3.7", "3.0")
- with open("infer/modules/train/preprocess.py", "w") as f:
- f.write(strr)
- elif self.has_mps():
- logger.info("No supported Nvidia GPU found")
- self.device = self.instead = "mps"
- self.is_half = False
- self.use_fp32_config()
- else:
- logger.info("No supported Nvidia GPU found")
- self.device = self.instead = "cpu"
- self.is_half = False
- self.use_fp32_config()
-
- if self.n_cpu == 0:
- self.n_cpu = cpu_count()
-
- if self.is_half:
- # 6G显存配置
- x_pad = 3
- x_query = 10
- x_center = 60
- x_max = 65
- else:
- # 5G显存配置
- x_pad = 1
- x_query = 6
- x_center = 38
- x_max = 41
-
- if self.gpu_mem is not None and self.gpu_mem <= 4:
- x_pad = 1
- x_query = 5
- x_center = 30
- x_max = 32
- if self.dml:
- logger.info("Use DirectML instead")
- if (
- os.path.exists(
- "runtime\Lib\site-packages\onnxruntime\capi\DirectML.dll"
- )
- == False
- ):
- try:
- os.rename(
- "runtime\Lib\site-packages\onnxruntime",
- "runtime\Lib\site-packages\onnxruntime-cuda",
- )
- except:
- pass
- try:
- os.rename(
- "runtime\Lib\site-packages\onnxruntime-dml",
- "runtime\Lib\site-packages\onnxruntime",
- )
- except:
- pass
- # if self.device != "cpu":
- import torch_directml
-
- self.device = torch_directml.device(torch_directml.default_device())
- self.is_half = False
- else:
- if self.instead:
- logger.info(f"Use {self.instead} instead")
- if (
- os.path.exists(
- "runtime\Lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll"
- )
- == False
- ):
- try:
- os.rename(
- "runtime\Lib\site-packages\onnxruntime",
- "runtime\Lib\site-packages\onnxruntime-dml",
- )
- except:
- pass
- try:
- os.rename(
- "runtime\Lib\site-packages\onnxruntime-cuda",
- "runtime\Lib\site-packages\onnxruntime",
- )
- except:
- pass
- return x_pad, x_query, x_center, x_max
diff --git a/core.py b/core.py
new file mode 100644
index 0000000000000000000000000000000000000000..3961b0aa5d971427126424e10b35af46215af68a
--- /dev/null
+++ b/core.py
@@ -0,0 +1,781 @@
+import os
+import sys
+import json
+import argparse
+import subprocess
+import spaces
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+
+from rvc.configs.config import Config
+
+from rvc.lib.tools.prerequisites_download import prequisites_download_pipeline
+
+from rvc.infer.infer import infer_pipeline
+
+from rvc.lib.tools.model_download import model_download_pipeline
+
+config = Config()
+current_script_directory = os.path.dirname(os.path.realpath(__file__))
+logs_path = os.path.join(current_script_directory, "logs")
+
+# Get TTS Voices
+with open(os.path.join("rvc", "lib", "tools", "tts_voices.json"), "r") as f:
+ voices_data = json.load(f)
+
+locales = list({voice["Locale"] for voice in voices_data})
+
+
+# Infer
+@spaces.GPU
+def run_infer_script(
+ f0up_key,
+ filter_radius,
+ index_rate,
+ rms_mix_rate,
+ protect,
+ hop_length,
+ f0method,
+ input_path,
+ output_path,
+ pth_path,
+ index_path,
+ split_audio,
+ f0autotune,
+ clean_audio,
+ clean_strength,
+ export_format,
+ embedder_model,
+ embedder_model_custom,
+ upscale_audio,
+):
+ f0autotune = "True" if str(f0autotune) == "True" else "False"
+ clean_audio = "True" if str(clean_audio) == "True" else "False"
+ upscale_audio = "True" if str(upscale_audio) == "True" else "False"
+ infer_pipeline(
+ f0up_key,
+ filter_radius,
+ index_rate,
+ rms_mix_rate,
+ protect,
+ hop_length,
+ f0method,
+ input_path,
+ output_path,
+ pth_path,
+ index_path,
+ split_audio,
+ f0autotune,
+ clean_audio,
+ clean_strength,
+ export_format,
+ embedder_model,
+ embedder_model_custom,
+ upscale_audio,
+ )
+ return f"File {input_path} inferred successfully.", output_path.replace(
+ ".wav", f".{export_format.lower()}"
+ )
+
+
+# Batch infer
+@spaces.GPU
+def run_batch_infer_script(
+ f0up_key,
+ filter_radius,
+ index_rate,
+ rms_mix_rate,
+ protect,
+ hop_length,
+ f0method,
+ input_folder,
+ output_folder,
+ pth_path,
+ index_path,
+ split_audio,
+ f0autotune,
+ clean_audio,
+ clean_strength,
+ export_format,
+ embedder_model,
+ embedder_model_custom,
+ upscale_audio,
+):
+ f0autotune = "True" if str(f0autotune) == "True" else "False"
+ clean_audio = "True" if str(clean_audio) == "True" else "False"
+ upscale_audio = "True" if str(upscale_audio) == "True" else "False"
+ audio_files = [
+ f for f in os.listdir(input_folder) if f.endswith((".mp3", ".wav", ".flac"))
+ ]
+ print(f"Detected {len(audio_files)} audio files for inference.")
+
+ for audio_file in audio_files:
+ if "_output" in audio_file:
+ pass
+ else:
+ input_path = os.path.join(input_folder, audio_file)
+ output_file_name = os.path.splitext(os.path.basename(audio_file))[0]
+ output_path = os.path.join(
+ output_folder,
+ f"{output_file_name}_output{os.path.splitext(audio_file)[1]}",
+ )
+ print(f"Inferring {input_path}...")
+
+ infer_pipeline(
+ f0up_key,
+ filter_radius,
+ index_rate,
+ rms_mix_rate,
+ protect,
+ hop_length,
+ f0method,
+ input_path,
+ output_path,
+ pth_path,
+ index_path,
+ split_audio,
+ f0autotune,
+ clean_audio,
+ clean_strength,
+ export_format,
+ embedder_model,
+ embedder_model_custom,
+ upscale_audio,
+ )
+
+ return f"Files from {input_folder} inferred successfully."
+
+
+# TTS
+@spaces.GPU
+def run_tts_script(
+ tts_text,
+ tts_voice,
+ tts_rate,
+ f0up_key,
+ filter_radius,
+ index_rate,
+ rms_mix_rate,
+ protect,
+ hop_length,
+ f0method,
+ output_tts_path,
+ output_rvc_path,
+ pth_path,
+ index_path,
+ split_audio,
+ f0autotune,
+ clean_audio,
+ clean_strength,
+ export_format,
+ embedder_model,
+ embedder_model_custom,
+ upscale_audio,
+):
+ f0autotune = "True" if str(f0autotune) == "True" else "False"
+ clean_audio = "True" if str(clean_audio) == "True" else "False"
+ upscale_audio = "True" if str(upscale_audio) == "True" else "False"
+ tts_script_path = os.path.join("rvc", "lib", "tools", "tts.py")
+
+ if os.path.exists(output_tts_path):
+ os.remove(output_tts_path)
+
+ command_tts = [
+ "python",
+ tts_script_path,
+ tts_text,
+ tts_voice,
+ str(tts_rate),
+ output_tts_path,
+ ]
+ subprocess.run(command_tts)
+
+ infer_pipeline(
+ f0up_key,
+ filter_radius,
+ index_rate,
+ rms_mix_rate,
+ protect,
+ hop_length,
+ f0method,
+ output_tts_path,
+ output_rvc_path,
+ pth_path,
+ index_path,
+ split_audio,
+ f0autotune,
+ clean_audio,
+ clean_strength,
+ export_format,
+ embedder_model,
+ embedder_model_custom,
+ upscale_audio,
+ )
+
+ return f"Text {tts_text} synthesized successfully.", output_rvc_path.replace(
+ ".wav", f".{export_format.lower()}"
+ )
+
+
+# Download
+def run_download_script(model_link):
+ model_download_pipeline(model_link)
+ return f"Model downloaded successfully."
+
+
+# Prerequisites
+def run_prerequisites_script(pretraineds_v1, pretraineds_v2, models, exe):
+ prequisites_download_pipeline(pretraineds_v1, pretraineds_v2, models, exe)
+ return "Prerequisites installed successfully."
+
+# Parse arguments
+def parse_arguments():
+ parser = argparse.ArgumentParser(
+ description="Run the main.py script with specific parameters."
+ )
+ subparsers = parser.add_subparsers(
+ title="subcommands", dest="mode", help="Choose a mode"
+ )
+
+ # Parser for 'infer' mode
+ infer_parser = subparsers.add_parser("infer", help="Run inference")
+ infer_parser.add_argument(
+ "--f0up_key",
+ type=str,
+ help="Value for f0up_key",
+ choices=[str(i) for i in range(-24, 25)],
+ default="0",
+ )
+ infer_parser.add_argument(
+ "--filter_radius",
+ type=str,
+ help="Value for filter_radius",
+ choices=[str(i) for i in range(11)],
+ default="3",
+ )
+ infer_parser.add_argument(
+ "--index_rate",
+ type=str,
+ help="Value for index_rate",
+ choices=[str(i / 10) for i in range(11)],
+ default="0.3",
+ )
+ infer_parser.add_argument(
+ "--rms_mix_rate",
+ type=str,
+ help="Value for rms_mix_rate",
+ choices=[str(i / 10) for i in range(11)],
+ default="1",
+ )
+ infer_parser.add_argument(
+ "--protect",
+ type=str,
+ help="Value for protect",
+ choices=[str(i / 10) for i in range(6)],
+ default="0.33",
+ )
+ infer_parser.add_argument(
+ "--hop_length",
+ type=str,
+ help="Value for hop_length",
+ choices=[str(i) for i in range(1, 513)],
+ default="128",
+ )
+ infer_parser.add_argument(
+ "--f0method",
+ type=str,
+ help="Value for f0method",
+ choices=[
+ "pm",
+ "harvest",
+ "dio",
+ "crepe",
+ "crepe-tiny",
+ "rmvpe",
+ "fcpe",
+ "hybrid[crepe+rmvpe]",
+ "hybrid[crepe+fcpe]",
+ "hybrid[rmvpe+fcpe]",
+ "hybrid[crepe+rmvpe+fcpe]",
+ ],
+ default="rmvpe",
+ )
+ infer_parser.add_argument("--input_path", type=str, help="Input path")
+ infer_parser.add_argument("--output_path", type=str, help="Output path")
+ infer_parser.add_argument("--pth_path", type=str, help="Path to the .pth file")
+ infer_parser.add_argument(
+ "--index_path",
+ type=str,
+ help="Path to the .index file",
+ )
+ infer_parser.add_argument(
+ "--split_audio",
+ type=str,
+ help="Enable split audio",
+ choices=["True", "False"],
+ default="False",
+ )
+ infer_parser.add_argument(
+ "--f0autotune",
+ type=str,
+ help="Enable autotune",
+ choices=["True", "False"],
+ default="False",
+ )
+ infer_parser.add_argument(
+ "--clean_audio",
+ type=str,
+ help="Enable clean audio",
+ choices=["True", "False"],
+ default="False",
+ )
+ infer_parser.add_argument(
+ "--clean_strength",
+ type=str,
+ help="Value for clean_strength",
+ choices=[str(i / 10) for i in range(11)],
+ default="0.7",
+ )
+ infer_parser.add_argument(
+ "--export_format",
+ type=str,
+ help="Export format",
+ choices=["WAV", "MP3", "FLAC", "OGG", "M4A"],
+ default="WAV",
+ )
+ infer_parser.add_argument(
+ "--embedder_model",
+ type=str,
+ help="Embedder model",
+ choices=["contentvec", "hubert", "custom"],
+ default="hubert",
+ )
+ infer_parser.add_argument(
+ "--embedder_model_custom",
+ type=str,
+ help="Custom Embedder model",
+ default=None,
+ )
+ infer_parser.add_argument(
+ "--upscale_audio",
+ type=str,
+ help="Enable audio upscaling",
+ choices=["True", "False"],
+ default="False",
+ )
+
+ # Parser for 'batch_infer' mode
+ batch_infer_parser = subparsers.add_parser(
+ "batch_infer", help="Run batch inference"
+ )
+ batch_infer_parser.add_argument(
+ "--f0up_key",
+ type=str,
+ help="Value for f0up_key",
+ choices=[str(i) for i in range(-24, 25)],
+ default="0",
+ )
+ batch_infer_parser.add_argument(
+ "--filter_radius",
+ type=str,
+ help="Value for filter_radius",
+ choices=[str(i) for i in range(11)],
+ default="3",
+ )
+ batch_infer_parser.add_argument(
+ "--index_rate",
+ type=str,
+ help="Value for index_rate",
+ choices=[str(i / 10) for i in range(11)],
+ default="0.3",
+ )
+ batch_infer_parser.add_argument(
+ "--rms_mix_rate",
+ type=str,
+ help="Value for rms_mix_rate",
+ choices=[str(i / 10) for i in range(11)],
+ default="1",
+ )
+ batch_infer_parser.add_argument(
+ "--protect",
+ type=str,
+ help="Value for protect",
+ choices=[str(i / 10) for i in range(6)],
+ default="0.33",
+ )
+ batch_infer_parser.add_argument(
+ "--hop_length",
+ type=str,
+ help="Value for hop_length",
+ choices=[str(i) for i in range(1, 513)],
+ default="128",
+ )
+ batch_infer_parser.add_argument(
+ "--f0method",
+ type=str,
+ help="Value for f0method",
+ choices=[
+ "pm",
+ "harvest",
+ "dio",
+ "crepe",
+ "crepe-tiny",
+ "rmvpe",
+ "fcpe",
+ "hybrid[crepe+rmvpe]",
+ "hybrid[crepe+fcpe]",
+ "hybrid[rmvpe+fcpe]",
+ "hybrid[crepe+rmvpe+fcpe]",
+ ],
+ default="rmvpe",
+ )
+ batch_infer_parser.add_argument("--input_folder", type=str, help="Input folder")
+ batch_infer_parser.add_argument("--output_folder", type=str, help="Output folder")
+ batch_infer_parser.add_argument(
+ "--pth_path", type=str, help="Path to the .pth file"
+ )
+ batch_infer_parser.add_argument(
+ "--index_path",
+ type=str,
+ help="Path to the .index file",
+ )
+ batch_infer_parser.add_argument(
+ "--split_audio",
+ type=str,
+ help="Enable split audio",
+ choices=["True", "False"],
+ default="False",
+ )
+ batch_infer_parser.add_argument(
+ "--f0autotune",
+ type=str,
+ help="Enable autotune",
+ choices=["True", "False"],
+ default="False",
+ )
+ batch_infer_parser.add_argument(
+ "--clean_audio",
+ type=str,
+ help="Enable clean audio",
+ choices=["True", "False"],
+ default="False",
+ )
+ batch_infer_parser.add_argument(
+ "--clean_strength",
+ type=str,
+ help="Value for clean_strength",
+ choices=[str(i / 10) for i in range(11)],
+ default="0.7",
+ )
+ batch_infer_parser.add_argument(
+ "--export_format",
+ type=str,
+ help="Export format",
+ choices=["WAV", "MP3", "FLAC", "OGG", "M4A"],
+ default="WAV",
+ )
+ batch_infer_parser.add_argument(
+ "--embedder_model",
+ type=str,
+ help="Embedder model",
+ choices=["contentvec", "hubert", "custom"],
+ default="hubert",
+ )
+ batch_infer_parser.add_argument(
+ "--embedder_model_custom",
+ type=str,
+ help="Custom Embedder model",
+ default=None,
+ )
+ batch_infer_parser.add_argument(
+ "--upscale_audio",
+ type=str,
+ help="Enable audio upscaling",
+ choices=["True", "False"],
+ default="False",
+ )
+
+ # Parser for 'tts' mode
+ tts_parser = subparsers.add_parser("tts", help="Run TTS")
+ tts_parser.add_argument(
+ "--tts_text",
+ type=str,
+ help="Text to be synthesized",
+ )
+ tts_parser.add_argument(
+ "--tts_voice",
+ type=str,
+ help="Voice to be used",
+ choices=locales,
+ )
+ tts_parser.add_argument(
+ "--tts_rate",
+ type=str,
+ help="Increase or decrease TTS speed",
+ choices=[str(i) for i in range(-100, 100)],
+ default="0",
+ )
+ tts_parser.add_argument(
+ "--f0up_key",
+ type=str,
+ help="Value for f0up_key",
+ choices=[str(i) for i in range(-24, 25)],
+ default="0",
+ )
+ tts_parser.add_argument(
+ "--filter_radius",
+ type=str,
+ help="Value for filter_radius",
+ choices=[str(i) for i in range(11)],
+ default="3",
+ )
+ tts_parser.add_argument(
+ "--index_rate",
+ type=str,
+ help="Value for index_rate",
+ choices=[str(i / 10) for i in range(11)],
+ default="0.3",
+ )
+ tts_parser.add_argument(
+ "--rms_mix_rate",
+ type=str,
+ help="Value for rms_mix_rate",
+ choices=[str(i / 10) for i in range(11)],
+ default="1",
+ )
+ tts_parser.add_argument(
+ "--protect",
+ type=str,
+ help="Value for protect",
+ choices=[str(i / 10) for i in range(6)],
+ default="0.33",
+ )
+ tts_parser.add_argument(
+ "--hop_length",
+ type=str,
+ help="Value for hop_length",
+ choices=[str(i) for i in range(1, 513)],
+ default="128",
+ )
+ tts_parser.add_argument(
+ "--f0method",
+ type=str,
+ help="Value for f0method",
+ choices=[
+ "pm",
+ "harvest",
+ "dio",
+ "crepe",
+ "crepe-tiny",
+ "rmvpe",
+ "fcpe",
+ "hybrid[crepe+rmvpe]",
+ "hybrid[crepe+fcpe]",
+ "hybrid[rmvpe+fcpe]",
+ "hybrid[crepe+rmvpe+fcpe]",
+ ],
+ default="rmvpe",
+ )
+ tts_parser.add_argument("--output_tts_path", type=str, help="Output tts path")
+ tts_parser.add_argument("--output_rvc_path", type=str, help="Output rvc path")
+ tts_parser.add_argument("--pth_path", type=str, help="Path to the .pth file")
+ tts_parser.add_argument(
+ "--index_path",
+ type=str,
+ help="Path to the .index file",
+ )
+ tts_parser.add_argument(
+ "--split_audio",
+ type=str,
+ help="Enable split audio",
+ choices=["True", "False"],
+ default="False",
+ )
+ tts_parser.add_argument(
+ "--f0autotune",
+ type=str,
+ help="Enable autotune",
+ choices=["True", "False"],
+ default="False",
+ )
+ tts_parser.add_argument(
+ "--clean_audio",
+ type=str,
+ help="Enable clean audio",
+ choices=["True", "False"],
+ default="False",
+ )
+ tts_parser.add_argument(
+ "--clean_strength",
+ type=str,
+ help="Value for clean_strength",
+ choices=[str(i / 10) for i in range(11)],
+ default="0.7",
+ )
+ tts_parser.add_argument(
+ "--export_format",
+ type=str,
+ help="Export format",
+ choices=["WAV", "MP3", "FLAC", "OGG", "M4A"],
+ default="WAV",
+ )
+ tts_parser.add_argument(
+ "--embedder_model",
+ type=str,
+ help="Embedder model",
+ choices=["contentvec", "hubert", "custom"],
+ default="hubert",
+ )
+ tts_parser.add_argument(
+ "--embedder_model_custom",
+ type=str,
+ help="Custom Embedder model",
+ default=None,
+ )
+ tts_parser.add_argument(
+ "--upscale_audio",
+ type=str,
+ help="Enable audio upscaling",
+ choices=["True", "False"],
+ default="False",
+ )
+
+ # Parser for 'download' mode
+ download_parser = subparsers.add_parser("download", help="Download models")
+ download_parser.add_argument(
+ "--model_link",
+ type=str,
+ help="Link of the model",
+ )
+
+ # Parser for 'prerequisites' mode
+ prerequisites_parser = subparsers.add_parser(
+ "prerequisites", help="Install prerequisites"
+ )
+ prerequisites_parser.add_argument(
+ "--pretraineds_v1",
+ type=str,
+ choices=["True", "False"],
+ default="True",
+ help="Download pretrained models for v1",
+ )
+ prerequisites_parser.add_argument(
+ "--pretraineds_v2",
+ type=str,
+ choices=["True", "False"],
+ default="True",
+ help="Download pretrained models for v2",
+ )
+ prerequisites_parser.add_argument(
+ "--models",
+ type=str,
+ choices=["True", "False"],
+ default="True",
+ help="Donwload models",
+ )
+ prerequisites_parser.add_argument(
+ "--exe",
+ type=str,
+ choices=["True", "False"],
+ default="True",
+ help="Download executables",
+ )
+
+ return parser.parse_args()
+
+
+def main():
+ if len(sys.argv) == 1:
+ print("Please run the script with '-h' for more information.")
+ sys.exit(1)
+
+ args = parse_arguments()
+
+ try:
+ if args.mode == "infer":
+ run_infer_script(
+ str(args.f0up_key),
+ str(args.filter_radius),
+ str(args.index_rate),
+ str(args.rms_mix_rate),
+ str(args.protect),
+ str(args.hop_length),
+ str(args.f0method),
+ str(args.input_path),
+ str(args.output_path),
+ str(args.pth_path),
+ str(args.index_path),
+ str(args.split_audio),
+ str(args.f0autotune),
+ str(args.clean_audio),
+ str(args.clean_strength),
+ str(args.export_format),
+ str(args.embedder_model),
+ str(args.embedder_model_custom),
+ str(args.upscale_audio),
+ )
+ elif args.mode == "batch_infer":
+ run_batch_infer_script(
+ str(args.f0up_key),
+ str(args.filter_radius),
+ str(args.index_rate),
+ str(args.rms_mix_rate),
+ str(args.protect),
+ str(args.hop_length),
+ str(args.f0method),
+ str(args.input_folder),
+ str(args.output_folder),
+ str(args.pth_path),
+ str(args.index_path),
+ str(args.split_audio),
+ str(args.f0autotune),
+ str(args.clean_audio),
+ str(args.clean_strength),
+ str(args.export_format),
+ str(args.embedder_model),
+ str(args.embedder_model_custom),
+ str(args.upscale_audio),
+ )
+ elif args.mode == "tts":
+ run_tts_script(
+ str(args.tts_text),
+ str(args.tts_voice),
+ str(args.tts_rate),
+ str(args.f0up_key),
+ str(args.filter_radius),
+ str(args.index_rate),
+ str(args.rms_mix_rate),
+ str(args.protect),
+ str(args.hop_length),
+ str(args.f0method),
+ str(args.output_tts_path),
+ str(args.output_rvc_path),
+ str(args.pth_path),
+ str(args.index_path),
+ str(args.split_audio),
+ str(args.f0autotune),
+ str(args.clean_audio),
+ str(args.clean_strength),
+ str(args.export_format),
+ str(args.embedder_model),
+ str(args.embedder_model_custom),
+ str(args.upscale_audio),
+ )
+ elif args.mode == "download":
+ run_download_script(
+ str(args.model_link),
+ )
+ elif args.mode == "prerequisites":
+ run_prerequisites_script(
+ str(args.pretraineds_v1),
+ str(args.pretraineds_v2),
+ str(args.models),
+ str(args.exe),
+ )
+ except Exception as error:
+ print(f"Error: {error}")
+
+
+if __name__ == "__main__":
+ main()
diff --git a/delete_models.py b/delete_models.py
new file mode 100644
index 0000000000000000000000000000000000000000..467ac6eb92be4e01e875a7259de2a644fb5826e0
--- /dev/null
+++ b/delete_models.py
@@ -0,0 +1,44 @@
+import sys
+import os
+import time
+import threading
+import shutil
+import logging
+
+now_dir = os.getcwd()
+sys.path.append(now_dir)
+
+logging.basicConfig(level=logging.DEBUG)
+
+def infinite_loop():
+ while True:
+ try:
+ models_folder = os.path.join(now_dir, "logs")
+
+ for element in os.listdir(models_folder):
+ element_route = os.path.join(models_folder, element)
+ if os.path.isdir(element_route) and element != "mute":
+ shutil.rmtree(element_route)
+ elif os.path.isfile(element_route):
+ os.remove(element_route)
+ except Exception as e:
+ logging.error(f"Error in models_folder loop: {e}")
+
+ try:
+ audios_folder = os.path.join(now_dir, "audios")
+
+ for element in os.listdir(audios_folder):
+ element_route = os.path.join(audios_folder, element)
+ if os.path.isfile(element_route):
+ os.remove(element_route)
+ except Exception as e:
+ logging.error(f"Error in audios_folder loop: {e}")
+
+ wait_time = 24 * 60 * 60
+ logging.info(f"Sleeping for {wait_time} seconds")
+ time.sleep(wait_time)
+
+def start_infinite_loop():
+ hilo_bucle = threading.Thread(target=infinite_loop)
+ hilo_bucle.daemon = True
+ hilo_bucle.start()
\ No newline at end of file
diff --git a/demucs/__init__.py b/demucs/__init__.py
deleted file mode 100644
index d4182e356427e1b05a79f8da641c70bb732514fa..0000000000000000000000000000000000000000
--- a/demucs/__init__.py
+++ /dev/null
@@ -1,7 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-__version__ = "2.0.3"
diff --git a/demucs/__main__.py b/demucs/__main__.py
deleted file mode 100644
index 5148f20623bdaa827777558844796ded1876d7d0..0000000000000000000000000000000000000000
--- a/demucs/__main__.py
+++ /dev/null
@@ -1,317 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import json
-import math
-import os
-import sys
-import time
-from dataclasses import dataclass, field
-
-import torch as th
-from torch import distributed, nn
-from torch.nn.parallel.distributed import DistributedDataParallel
-
-from .augment import FlipChannels, FlipSign, Remix, Scale, Shift
-from .compressed import get_compressed_datasets
-from .model import Demucs
-from .parser import get_name, get_parser
-from .raw import Rawset
-from .repitch import RepitchedWrapper
-from .pretrained import load_pretrained, SOURCES
-from .tasnet import ConvTasNet
-from .test import evaluate
-from .train import train_model, validate_model
-from .utils import (human_seconds, load_model, save_model, get_state,
- save_state, sizeof_fmt, get_quantizer)
-from .wav import get_wav_datasets, get_musdb_wav_datasets
-
-
-@dataclass
-class SavedState:
- metrics: list = field(default_factory=list)
- last_state: dict = None
- best_state: dict = None
- optimizer: dict = None
-
-
-def main():
- parser = get_parser()
- args = parser.parse_args()
- name = get_name(parser, args)
- print(f"Experiment {name}")
-
- if args.musdb is None and args.rank == 0:
- print(
- "You must provide the path to the MusDB dataset with the --musdb flag. "
- "To download the MusDB dataset, see https://sigsep.github.io/datasets/musdb.html.",
- file=sys.stderr)
- sys.exit(1)
-
- eval_folder = args.evals / name
- eval_folder.mkdir(exist_ok=True, parents=True)
- args.logs.mkdir(exist_ok=True)
- metrics_path = args.logs / f"{name}.json"
- eval_folder.mkdir(exist_ok=True, parents=True)
- args.checkpoints.mkdir(exist_ok=True, parents=True)
- args.models.mkdir(exist_ok=True, parents=True)
-
- if args.device is None:
- device = "cpu"
- if th.cuda.is_available():
- device = "cuda"
- else:
- device = args.device
-
- th.manual_seed(args.seed)
- # Prevents too many threads to be started when running `museval` as it can be quite
- # inefficient on NUMA architectures.
- os.environ["OMP_NUM_THREADS"] = "1"
- os.environ["MKL_NUM_THREADS"] = "1"
-
- if args.world_size > 1:
- if device != "cuda" and args.rank == 0:
- print("Error: distributed training is only available with cuda device", file=sys.stderr)
- sys.exit(1)
- th.cuda.set_device(args.rank % th.cuda.device_count())
- distributed.init_process_group(backend="nccl",
- init_method="tcp://" + args.master,
- rank=args.rank,
- world_size=args.world_size)
-
- checkpoint = args.checkpoints / f"{name}.th"
- checkpoint_tmp = args.checkpoints / f"{name}.th.tmp"
- if args.restart and checkpoint.exists() and args.rank == 0:
- checkpoint.unlink()
-
- if args.test or args.test_pretrained:
- args.epochs = 1
- args.repeat = 0
- if args.test:
- model = load_model(args.models / args.test)
- else:
- model = load_pretrained(args.test_pretrained)
- elif args.tasnet:
- model = ConvTasNet(audio_channels=args.audio_channels,
- samplerate=args.samplerate, X=args.X,
- segment_length=4 * args.samples,
- sources=SOURCES)
- else:
- model = Demucs(
- audio_channels=args.audio_channels,
- channels=args.channels,
- context=args.context,
- depth=args.depth,
- glu=args.glu,
- growth=args.growth,
- kernel_size=args.kernel_size,
- lstm_layers=args.lstm_layers,
- rescale=args.rescale,
- rewrite=args.rewrite,
- stride=args.conv_stride,
- resample=args.resample,
- normalize=args.normalize,
- samplerate=args.samplerate,
- segment_length=4 * args.samples,
- sources=SOURCES,
- )
- model.to(device)
- if args.init:
- model.load_state_dict(load_pretrained(args.init).state_dict())
-
- if args.show:
- print(model)
- size = sizeof_fmt(4 * sum(p.numel() for p in model.parameters()))
- print(f"Model size {size}")
- return
-
- try:
- saved = th.load(checkpoint, map_location='cpu')
- except IOError:
- saved = SavedState()
-
- optimizer = th.optim.Adam(model.parameters(), lr=args.lr)
-
- quantizer = None
- quantizer = get_quantizer(model, args, optimizer)
-
- if saved.last_state is not None:
- model.load_state_dict(saved.last_state, strict=False)
- if saved.optimizer is not None:
- optimizer.load_state_dict(saved.optimizer)
-
- model_name = f"{name}.th"
- if args.save_model:
- if args.rank == 0:
- model.to("cpu")
- model.load_state_dict(saved.best_state)
- save_model(model, quantizer, args, args.models / model_name)
- return
- elif args.save_state:
- model_name = f"{args.save_state}.th"
- if args.rank == 0:
- model.to("cpu")
- model.load_state_dict(saved.best_state)
- state = get_state(model, quantizer)
- save_state(state, args.models / model_name)
- return
-
- if args.rank == 0:
- done = args.logs / f"{name}.done"
- if done.exists():
- done.unlink()
-
- augment = [Shift(args.data_stride)]
- if args.augment:
- augment += [FlipSign(), FlipChannels(), Scale(),
- Remix(group_size=args.remix_group_size)]
- augment = nn.Sequential(*augment).to(device)
- print("Agumentation pipeline:", augment)
-
- if args.mse:
- criterion = nn.MSELoss()
- else:
- criterion = nn.L1Loss()
-
- # Setting number of samples so that all convolution windows are full.
- # Prevents hard to debug mistake with the prediction being shifted compared
- # to the input mixture.
- samples = model.valid_length(args.samples)
- print(f"Number of training samples adjusted to {samples}")
- samples = samples + args.data_stride
- if args.repitch:
- # We need a bit more audio samples, to account for potential
- # tempo change.
- samples = math.ceil(samples / (1 - 0.01 * args.max_tempo))
-
- args.metadata.mkdir(exist_ok=True, parents=True)
- if args.raw:
- train_set = Rawset(args.raw / "train",
- samples=samples,
- channels=args.audio_channels,
- streams=range(1, len(model.sources) + 1),
- stride=args.data_stride)
-
- valid_set = Rawset(args.raw / "valid", channels=args.audio_channels)
- elif args.wav:
- train_set, valid_set = get_wav_datasets(args, samples, model.sources)
- elif args.is_wav:
- train_set, valid_set = get_musdb_wav_datasets(args, samples, model.sources)
- else:
- train_set, valid_set = get_compressed_datasets(args, samples)
-
- if args.repitch:
- train_set = RepitchedWrapper(
- train_set,
- proba=args.repitch,
- max_tempo=args.max_tempo)
-
- best_loss = float("inf")
- for epoch, metrics in enumerate(saved.metrics):
- print(f"Epoch {epoch:03d}: "
- f"train={metrics['train']:.8f} "
- f"valid={metrics['valid']:.8f} "
- f"best={metrics['best']:.4f} "
- f"ms={metrics.get('true_model_size', 0):.2f}MB "
- f"cms={metrics.get('compressed_model_size', 0):.2f}MB "
- f"duration={human_seconds(metrics['duration'])}")
- best_loss = metrics['best']
-
- if args.world_size > 1:
- dmodel = DistributedDataParallel(model,
- device_ids=[th.cuda.current_device()],
- output_device=th.cuda.current_device())
- else:
- dmodel = model
-
- for epoch in range(len(saved.metrics), args.epochs):
- begin = time.time()
- model.train()
- train_loss, model_size = train_model(
- epoch, train_set, dmodel, criterion, optimizer, augment,
- quantizer=quantizer,
- batch_size=args.batch_size,
- device=device,
- repeat=args.repeat,
- seed=args.seed,
- diffq=args.diffq,
- workers=args.workers,
- world_size=args.world_size)
- model.eval()
- valid_loss = validate_model(
- epoch, valid_set, model, criterion,
- device=device,
- rank=args.rank,
- split=args.split_valid,
- overlap=args.overlap,
- world_size=args.world_size)
-
- ms = 0
- cms = 0
- if quantizer and args.rank == 0:
- ms = quantizer.true_model_size()
- cms = quantizer.compressed_model_size(num_workers=min(40, args.world_size * 10))
-
- duration = time.time() - begin
- if valid_loss < best_loss and ms <= args.ms_target:
- best_loss = valid_loss
- saved.best_state = {
- key: value.to("cpu").clone()
- for key, value in model.state_dict().items()
- }
-
- saved.metrics.append({
- "train": train_loss,
- "valid": valid_loss,
- "best": best_loss,
- "duration": duration,
- "model_size": model_size,
- "true_model_size": ms,
- "compressed_model_size": cms,
- })
- if args.rank == 0:
- json.dump(saved.metrics, open(metrics_path, "w"))
-
- saved.last_state = model.state_dict()
- saved.optimizer = optimizer.state_dict()
- if args.rank == 0 and not args.test:
- th.save(saved, checkpoint_tmp)
- checkpoint_tmp.rename(checkpoint)
-
- print(f"Epoch {epoch:03d}: "
- f"train={train_loss:.8f} valid={valid_loss:.8f} best={best_loss:.4f} ms={ms:.2f}MB "
- f"cms={cms:.2f}MB "
- f"duration={human_seconds(duration)}")
-
- if args.world_size > 1:
- distributed.barrier()
-
- del dmodel
- model.load_state_dict(saved.best_state)
- if args.eval_cpu:
- device = "cpu"
- model.to(device)
- model.eval()
- evaluate(model, args.musdb, eval_folder,
- is_wav=args.is_wav,
- rank=args.rank,
- world_size=args.world_size,
- device=device,
- save=args.save,
- split=args.split_valid,
- shifts=args.shifts,
- overlap=args.overlap,
- workers=args.eval_workers)
- model.to("cpu")
- if args.rank == 0:
- if not (args.test or args.test_pretrained):
- save_model(model, quantizer, args, args.models / model_name)
- print("done")
- done.write_text("done")
-
-
-if __name__ == "__main__":
- main()
diff --git a/demucs/audio.py b/demucs/audio.py
deleted file mode 100644
index b29f156e4afb5fbda32c35777022caeadf50d711..0000000000000000000000000000000000000000
--- a/demucs/audio.py
+++ /dev/null
@@ -1,172 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-import json
-import subprocess as sp
-from pathlib import Path
-
-import julius
-import numpy as np
-import torch
-
-from .utils import temp_filenames
-
-
-def _read_info(path):
- stdout_data = sp.check_output([
- 'ffprobe', "-loglevel", "panic",
- str(path), '-print_format', 'json', '-show_format', '-show_streams'
- ])
- return json.loads(stdout_data.decode('utf-8'))
-
-
-class AudioFile:
- """
- Allows to read audio from any format supported by ffmpeg, as well as resampling or
- converting to mono on the fly. See :method:`read` for more details.
- """
- def __init__(self, path: Path):
- self.path = Path(path)
- self._info = None
-
- def __repr__(self):
- features = [("path", self.path)]
- features.append(("samplerate", self.samplerate()))
- features.append(("channels", self.channels()))
- features.append(("streams", len(self)))
- features_str = ", ".join(f"{name}={value}" for name, value in features)
- return f"AudioFile({features_str})"
-
- @property
- def info(self):
- if self._info is None:
- self._info = _read_info(self.path)
- return self._info
-
- @property
- def duration(self):
- return float(self.info['format']['duration'])
-
- @property
- def _audio_streams(self):
- return [
- index for index, stream in enumerate(self.info["streams"])
- if stream["codec_type"] == "audio"
- ]
-
- def __len__(self):
- return len(self._audio_streams)
-
- def channels(self, stream=0):
- return int(self.info['streams'][self._audio_streams[stream]]['channels'])
-
- def samplerate(self, stream=0):
- return int(self.info['streams'][self._audio_streams[stream]]['sample_rate'])
-
- def read(self,
- seek_time=None,
- duration=None,
- streams=slice(None),
- samplerate=None,
- channels=None,
- temp_folder=None):
- """
- Slightly more efficient implementation than stempeg,
- in particular, this will extract all stems at once
- rather than having to loop over one file multiple times
- for each stream.
-
- Args:
- seek_time (float): seek time in seconds or None if no seeking is needed.
- duration (float): duration in seconds to extract or None to extract until the end.
- streams (slice, int or list): streams to extract, can be a single int, a list or
- a slice. If it is a slice or list, the output will be of size [S, C, T]
- with S the number of streams, C the number of channels and T the number of samples.
- If it is an int, the output will be [C, T].
- samplerate (int): if provided, will resample on the fly. If None, no resampling will
- be done. Original sampling rate can be obtained with :method:`samplerate`.
- channels (int): if 1, will convert to mono. We do not rely on ffmpeg for that
- as ffmpeg automatically scale by +3dB to conserve volume when playing on speakers.
- See https://sound.stackexchange.com/a/42710.
- Our definition of mono is simply the average of the two channels. Any other
- value will be ignored.
- temp_folder (str or Path or None): temporary folder to use for decoding.
-
-
- """
- streams = np.array(range(len(self)))[streams]
- single = not isinstance(streams, np.ndarray)
- if single:
- streams = [streams]
-
- if duration is None:
- target_size = None
- query_duration = None
- else:
- target_size = int((samplerate or self.samplerate()) * duration)
- query_duration = float((target_size + 1) / (samplerate or self.samplerate()))
-
- with temp_filenames(len(streams)) as filenames:
- command = ['ffmpeg', '-y']
- command += ['-loglevel', 'panic']
- if seek_time:
- command += ['-ss', str(seek_time)]
- command += ['-i', str(self.path)]
- for stream, filename in zip(streams, filenames):
- command += ['-map', f'0:{self._audio_streams[stream]}']
- if query_duration is not None:
- command += ['-t', str(query_duration)]
- command += ['-threads', '1']
- command += ['-f', 'f32le']
- if samplerate is not None:
- command += ['-ar', str(samplerate)]
- command += [filename]
-
- sp.run(command, check=True)
- wavs = []
- for filename in filenames:
- wav = np.fromfile(filename, dtype=np.float32)
- wav = torch.from_numpy(wav)
- wav = wav.view(-1, self.channels()).t()
- if channels is not None:
- wav = convert_audio_channels(wav, channels)
- if target_size is not None:
- wav = wav[..., :target_size]
- wavs.append(wav)
- wav = torch.stack(wavs, dim=0)
- if single:
- wav = wav[0]
- return wav
-
-
-def convert_audio_channels(wav, channels=2):
- """Convert audio to the given number of channels."""
- *shape, src_channels, length = wav.shape
- if src_channels == channels:
- pass
- elif channels == 1:
- # Case 1:
- # The caller asked 1-channel audio, but the stream have multiple
- # channels, downmix all channels.
- wav = wav.mean(dim=-2, keepdim=True)
- elif src_channels == 1:
- # Case 2:
- # The caller asked for multiple channels, but the input file have
- # one single channel, replicate the audio over all channels.
- wav = wav.expand(*shape, channels, length)
- elif src_channels >= channels:
- # Case 3:
- # The caller asked for multiple channels, and the input file have
- # more channels than requested. In that case return the first channels.
- wav = wav[..., :channels, :]
- else:
- # Case 4: What is a reasonable choice here?
- raise ValueError('The audio file has less channels than requested but is not mono.')
- return wav
-
-
-def convert_audio(wav, from_samplerate, to_samplerate, channels):
- wav = convert_audio_channels(wav, channels)
- return julius.resample_frac(wav, from_samplerate, to_samplerate)
diff --git a/demucs/augment.py b/demucs/augment.py
deleted file mode 100644
index bb36d3298d89470f306316322e7587187819c94b..0000000000000000000000000000000000000000
--- a/demucs/augment.py
+++ /dev/null
@@ -1,106 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import random
-import torch as th
-from torch import nn
-
-
-class Shift(nn.Module):
- """
- Randomly shift audio in time by up to `shift` samples.
- """
- def __init__(self, shift=8192):
- super().__init__()
- self.shift = shift
-
- def forward(self, wav):
- batch, sources, channels, time = wav.size()
- length = time - self.shift
- if self.shift > 0:
- if not self.training:
- wav = wav[..., :length]
- else:
- offsets = th.randint(self.shift, [batch, sources, 1, 1], device=wav.device)
- offsets = offsets.expand(-1, -1, channels, -1)
- indexes = th.arange(length, device=wav.device)
- wav = wav.gather(3, indexes + offsets)
- return wav
-
-
-class FlipChannels(nn.Module):
- """
- Flip left-right channels.
- """
- def forward(self, wav):
- batch, sources, channels, time = wav.size()
- if self.training and wav.size(2) == 2:
- left = th.randint(2, (batch, sources, 1, 1), device=wav.device)
- left = left.expand(-1, -1, -1, time)
- right = 1 - left
- wav = th.cat([wav.gather(2, left), wav.gather(2, right)], dim=2)
- return wav
-
-
-class FlipSign(nn.Module):
- """
- Random sign flip.
- """
- def forward(self, wav):
- batch, sources, channels, time = wav.size()
- if self.training:
- signs = th.randint(2, (batch, sources, 1, 1), device=wav.device, dtype=th.float32)
- wav = wav * (2 * signs - 1)
- return wav
-
-
-class Remix(nn.Module):
- """
- Shuffle sources to make new mixes.
- """
- def __init__(self, group_size=4):
- """
- Shuffle sources within one batch.
- Each batch is divided into groups of size `group_size` and shuffling is done within
- each group separatly. This allow to keep the same probability distribution no matter
- the number of GPUs. Without this grouping, using more GPUs would lead to a higher
- probability of keeping two sources from the same track together which can impact
- performance.
- """
- super().__init__()
- self.group_size = group_size
-
- def forward(self, wav):
- batch, streams, channels, time = wav.size()
- device = wav.device
-
- if self.training:
- group_size = self.group_size or batch
- if batch % group_size != 0:
- raise ValueError(f"Batch size {batch} must be divisible by group size {group_size}")
- groups = batch // group_size
- wav = wav.view(groups, group_size, streams, channels, time)
- permutations = th.argsort(th.rand(groups, group_size, streams, 1, 1, device=device),
- dim=1)
- wav = wav.gather(1, permutations.expand(-1, -1, -1, channels, time))
- wav = wav.view(batch, streams, channels, time)
- return wav
-
-
-class Scale(nn.Module):
- def __init__(self, proba=1., min=0.25, max=1.25):
- super().__init__()
- self.proba = proba
- self.min = min
- self.max = max
-
- def forward(self, wav):
- batch, streams, channels, time = wav.size()
- device = wav.device
- if self.training and random.random() < self.proba:
- scales = th.empty(batch, streams, 1, 1, device=device).uniform_(self.min, self.max)
- wav *= scales
- return wav
diff --git a/demucs/compressed.py b/demucs/compressed.py
deleted file mode 100644
index eb8fbb75463ba71ca86729b22baebf24598ade57..0000000000000000000000000000000000000000
--- a/demucs/compressed.py
+++ /dev/null
@@ -1,115 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import json
-from fractions import Fraction
-from concurrent import futures
-
-import musdb
-from torch import distributed
-
-from .audio import AudioFile
-
-
-def get_musdb_tracks(root, *args, **kwargs):
- mus = musdb.DB(root, *args, **kwargs)
- return {track.name: track.path for track in mus}
-
-
-class StemsSet:
- def __init__(self, tracks, metadata, duration=None, stride=1,
- samplerate=44100, channels=2, streams=slice(None)):
-
- self.metadata = []
- for name, path in tracks.items():
- meta = dict(metadata[name])
- meta["path"] = path
- meta["name"] = name
- self.metadata.append(meta)
- if duration is not None and meta["duration"] < duration:
- raise ValueError(f"Track {name} duration is too small {meta['duration']}")
- self.metadata.sort(key=lambda x: x["name"])
- self.duration = duration
- self.stride = stride
- self.channels = channels
- self.samplerate = samplerate
- self.streams = streams
-
- def __len__(self):
- return sum(self._examples_count(m) for m in self.metadata)
-
- def _examples_count(self, meta):
- if self.duration is None:
- return 1
- else:
- return int((meta["duration"] - self.duration) // self.stride + 1)
-
- def track_metadata(self, index):
- for meta in self.metadata:
- examples = self._examples_count(meta)
- if index >= examples:
- index -= examples
- continue
- return meta
-
- def __getitem__(self, index):
- for meta in self.metadata:
- examples = self._examples_count(meta)
- if index >= examples:
- index -= examples
- continue
- streams = AudioFile(meta["path"]).read(seek_time=index * self.stride,
- duration=self.duration,
- channels=self.channels,
- samplerate=self.samplerate,
- streams=self.streams)
- return (streams - meta["mean"]) / meta["std"]
-
-
-def _get_track_metadata(path):
- # use mono at 44kHz as reference. For any other settings data won't be perfectly
- # normalized but it should be good enough.
- audio = AudioFile(path)
- mix = audio.read(streams=0, channels=1, samplerate=44100)
- return {"duration": audio.duration, "std": mix.std().item(), "mean": mix.mean().item()}
-
-
-def _build_metadata(tracks, workers=10):
- pendings = []
- with futures.ProcessPoolExecutor(workers) as pool:
- for name, path in tracks.items():
- pendings.append((name, pool.submit(_get_track_metadata, path)))
- return {name: p.result() for name, p in pendings}
-
-
-def _build_musdb_metadata(path, musdb, workers):
- tracks = get_musdb_tracks(musdb)
- metadata = _build_metadata(tracks, workers)
- path.parent.mkdir(exist_ok=True, parents=True)
- json.dump(metadata, open(path, "w"))
-
-
-def get_compressed_datasets(args, samples):
- metadata_file = args.metadata / "musdb.json"
- if not metadata_file.is_file() and args.rank == 0:
- _build_musdb_metadata(metadata_file, args.musdb, args.workers)
- if args.world_size > 1:
- distributed.barrier()
- metadata = json.load(open(metadata_file))
- duration = Fraction(samples, args.samplerate)
- stride = Fraction(args.data_stride, args.samplerate)
- train_set = StemsSet(get_musdb_tracks(args.musdb, subsets=["train"], split="train"),
- metadata,
- duration=duration,
- stride=stride,
- streams=slice(1, None),
- samplerate=args.samplerate,
- channels=args.audio_channels)
- valid_set = StemsSet(get_musdb_tracks(args.musdb, subsets=["train"], split="valid"),
- metadata,
- samplerate=args.samplerate,
- channels=args.audio_channels)
- return train_set, valid_set
diff --git a/demucs/model.py b/demucs/model.py
deleted file mode 100644
index e9d932f4d014f7b95b394d2e24ed5edc379ded8d..0000000000000000000000000000000000000000
--- a/demucs/model.py
+++ /dev/null
@@ -1,202 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import math
-
-import julius
-from torch import nn
-
-from .utils import capture_init, center_trim
-
-
-class BLSTM(nn.Module):
- def __init__(self, dim, layers=1):
- super().__init__()
- self.lstm = nn.LSTM(bidirectional=True, num_layers=layers, hidden_size=dim, input_size=dim)
- self.linear = nn.Linear(2 * dim, dim)
-
- def forward(self, x):
- x = x.permute(2, 0, 1)
- x = self.lstm(x)[0]
- x = self.linear(x)
- x = x.permute(1, 2, 0)
- return x
-
-
-def rescale_conv(conv, reference):
- std = conv.weight.std().detach()
- scale = (std / reference)**0.5
- conv.weight.data /= scale
- if conv.bias is not None:
- conv.bias.data /= scale
-
-
-def rescale_module(module, reference):
- for sub in module.modules():
- if isinstance(sub, (nn.Conv1d, nn.ConvTranspose1d)):
- rescale_conv(sub, reference)
-
-
-class Demucs(nn.Module):
- @capture_init
- def __init__(self,
- sources,
- audio_channels=2,
- channels=64,
- depth=6,
- rewrite=True,
- glu=True,
- rescale=0.1,
- resample=True,
- kernel_size=8,
- stride=4,
- growth=2.,
- lstm_layers=2,
- context=3,
- normalize=False,
- samplerate=44100,
- segment_length=4 * 10 * 44100):
- """
- Args:
- sources (list[str]): list of source names
- audio_channels (int): stereo or mono
- channels (int): first convolution channels
- depth (int): number of encoder/decoder layers
- rewrite (bool): add 1x1 convolution to each encoder layer
- and a convolution to each decoder layer.
- For the decoder layer, `context` gives the kernel size.
- glu (bool): use glu instead of ReLU
- resample_input (bool): upsample x2 the input and downsample /2 the output.
- rescale (int): rescale initial weights of convolutions
- to get their standard deviation closer to `rescale`
- kernel_size (int): kernel size for convolutions
- stride (int): stride for convolutions
- growth (float): multiply (resp divide) number of channels by that
- for each layer of the encoder (resp decoder)
- lstm_layers (int): number of lstm layers, 0 = no lstm
- context (int): kernel size of the convolution in the
- decoder before the transposed convolution. If > 1,
- will provide some context from neighboring time
- steps.
- samplerate (int): stored as meta information for easing
- future evaluations of the model.
- segment_length (int): stored as meta information for easing
- future evaluations of the model. Length of the segments on which
- the model was trained.
- """
-
- super().__init__()
- self.audio_channels = audio_channels
- self.sources = sources
- self.kernel_size = kernel_size
- self.context = context
- self.stride = stride
- self.depth = depth
- self.resample = resample
- self.channels = channels
- self.normalize = normalize
- self.samplerate = samplerate
- self.segment_length = segment_length
-
- self.encoder = nn.ModuleList()
- self.decoder = nn.ModuleList()
-
- if glu:
- activation = nn.GLU(dim=1)
- ch_scale = 2
- else:
- activation = nn.ReLU()
- ch_scale = 1
- in_channels = audio_channels
- for index in range(depth):
- encode = []
- encode += [nn.Conv1d(in_channels, channels, kernel_size, stride), nn.ReLU()]
- if rewrite:
- encode += [nn.Conv1d(channels, ch_scale * channels, 1), activation]
- self.encoder.append(nn.Sequential(*encode))
-
- decode = []
- if index > 0:
- out_channels = in_channels
- else:
- out_channels = len(self.sources) * audio_channels
- if rewrite:
- decode += [nn.Conv1d(channels, ch_scale * channels, context), activation]
- decode += [nn.ConvTranspose1d(channels, out_channels, kernel_size, stride)]
- if index > 0:
- decode.append(nn.ReLU())
- self.decoder.insert(0, nn.Sequential(*decode))
- in_channels = channels
- channels = int(growth * channels)
-
- channels = in_channels
-
- if lstm_layers:
- self.lstm = BLSTM(channels, lstm_layers)
- else:
- self.lstm = None
-
- if rescale:
- rescale_module(self, reference=rescale)
-
- def valid_length(self, length):
- """
- Return the nearest valid length to use with the model so that
- there is no time steps left over in a convolutions, e.g. for all
- layers, size of the input - kernel_size % stride = 0.
-
- If the mixture has a valid length, the estimated sources
- will have exactly the same length when context = 1. If context > 1,
- the two signals can be center trimmed to match.
-
- For training, extracts should have a valid length.For evaluation
- on full tracks we recommend passing `pad = True` to :method:`forward`.
- """
- if self.resample:
- length *= 2
- for _ in range(self.depth):
- length = math.ceil((length - self.kernel_size) / self.stride) + 1
- length = max(1, length)
- length += self.context - 1
- for _ in range(self.depth):
- length = (length - 1) * self.stride + self.kernel_size
-
- if self.resample:
- length = math.ceil(length / 2)
- return int(length)
-
- def forward(self, mix):
- x = mix
-
- if self.normalize:
- mono = mix.mean(dim=1, keepdim=True)
- mean = mono.mean(dim=-1, keepdim=True)
- std = mono.std(dim=-1, keepdim=True)
- else:
- mean = 0
- std = 1
-
- x = (x - mean) / (1e-5 + std)
-
- if self.resample:
- x = julius.resample_frac(x, 1, 2)
-
- saved = []
- for encode in self.encoder:
- x = encode(x)
- saved.append(x)
- if self.lstm:
- x = self.lstm(x)
- for decode in self.decoder:
- skip = center_trim(saved.pop(-1), x)
- x = x + skip
- x = decode(x)
-
- if self.resample:
- x = julius.resample_frac(x, 2, 1)
- x = x * std + mean
- x = x.view(x.size(0), len(self.sources), self.audio_channels, x.size(-1))
- return x
diff --git a/demucs/parser.py b/demucs/parser.py
deleted file mode 100644
index 4e8a19cf976e3c6dfe411da64b8dce3e9a4548e0..0000000000000000000000000000000000000000
--- a/demucs/parser.py
+++ /dev/null
@@ -1,244 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import argparse
-import os
-from pathlib import Path
-
-
-def get_parser():
- parser = argparse.ArgumentParser("demucs", description="Train and evaluate Demucs.")
- default_raw = None
- default_musdb = None
- if 'DEMUCS_RAW' in os.environ:
- default_raw = Path(os.environ['DEMUCS_RAW'])
- if 'DEMUCS_MUSDB' in os.environ:
- default_musdb = Path(os.environ['DEMUCS_MUSDB'])
- parser.add_argument(
- "--raw",
- type=Path,
- default=default_raw,
- help="Path to raw audio, can be faster, see python3 -m demucs.raw to extract.")
- parser.add_argument("--no_raw", action="store_const", const=None, dest="raw")
- parser.add_argument("-m",
- "--musdb",
- type=Path,
- default=default_musdb,
- help="Path to musdb root")
- parser.add_argument("--is_wav", action="store_true",
- help="Indicate that the MusDB dataset is in wav format (i.e. MusDB-HQ).")
- parser.add_argument("--metadata", type=Path, default=Path("metadata/"),
- help="Folder where metadata information is stored.")
- parser.add_argument("--wav", type=Path,
- help="Path to a wav dataset. This should contain a 'train' and a 'valid' "
- "subfolder.")
- parser.add_argument("--samplerate", type=int, default=44100)
- parser.add_argument("--audio_channels", type=int, default=2)
- parser.add_argument("--samples",
- default=44100 * 10,
- type=int,
- help="number of samples to feed in")
- parser.add_argument("--data_stride",
- default=44100,
- type=int,
- help="Stride for chunks, shorter = longer epochs")
- parser.add_argument("-w", "--workers", default=10, type=int, help="Loader workers")
- parser.add_argument("--eval_workers", default=2, type=int, help="Final evaluation workers")
- parser.add_argument("-d",
- "--device",
- help="Device to train on, default is cuda if available else cpu")
- parser.add_argument("--eval_cpu", action="store_true", help="Eval on test will be run on cpu.")
- parser.add_argument("--dummy", help="Dummy parameter, useful to create a new checkpoint file")
- parser.add_argument("--test", help="Just run the test pipeline + one validation. "
- "This should be a filename relative to the models/ folder.")
- parser.add_argument("--test_pretrained", help="Just run the test pipeline + one validation, "
- "on a pretrained model. ")
-
- parser.add_argument("--rank", default=0, type=int)
- parser.add_argument("--world_size", default=1, type=int)
- parser.add_argument("--master")
-
- parser.add_argument("--checkpoints",
- type=Path,
- default=Path("checkpoints"),
- help="Folder where to store checkpoints etc")
- parser.add_argument("--evals",
- type=Path,
- default=Path("evals"),
- help="Folder where to store evals and waveforms")
- parser.add_argument("--save",
- action="store_true",
- help="Save estimated for the test set waveforms")
- parser.add_argument("--logs",
- type=Path,
- default=Path("logs"),
- help="Folder where to store logs")
- parser.add_argument("--models",
- type=Path,
- default=Path("models"),
- help="Folder where to store trained models")
- parser.add_argument("-R",
- "--restart",
- action='store_true',
- help='Restart training, ignoring previous run')
-
- parser.add_argument("--seed", type=int, default=42)
- parser.add_argument("-e", "--epochs", type=int, default=180, help="Number of epochs")
- parser.add_argument("-r",
- "--repeat",
- type=int,
- default=2,
- help="Repeat the train set, longer epochs")
- parser.add_argument("-b", "--batch_size", type=int, default=64)
- parser.add_argument("--lr", type=float, default=3e-4)
- parser.add_argument("--mse", action="store_true", help="Use MSE instead of L1")
- parser.add_argument("--init", help="Initialize from a pre-trained model.")
-
- # Augmentation options
- parser.add_argument("--no_augment",
- action="store_false",
- dest="augment",
- default=True,
- help="No basic data augmentation.")
- parser.add_argument("--repitch", type=float, default=0.2,
- help="Probability to do tempo/pitch change")
- parser.add_argument("--max_tempo", type=float, default=12,
- help="Maximum relative tempo change in %% when using repitch.")
-
- parser.add_argument("--remix_group_size",
- type=int,
- default=4,
- help="Shuffle sources using group of this size. Useful to somewhat "
- "replicate multi-gpu training "
- "on less GPUs.")
- parser.add_argument("--shifts",
- type=int,
- default=10,
- help="Number of random shifts used for the shift trick.")
- parser.add_argument("--overlap",
- type=float,
- default=0.25,
- help="Overlap when --split_valid is passed.")
-
- # See model.py for doc
- parser.add_argument("--growth",
- type=float,
- default=2.,
- help="Number of channels between two layers will increase by this factor")
- parser.add_argument("--depth",
- type=int,
- default=6,
- help="Number of layers for the encoder and decoder")
- parser.add_argument("--lstm_layers", type=int, default=2, help="Number of layers for the LSTM")
- parser.add_argument("--channels",
- type=int,
- default=64,
- help="Number of channels for the first encoder layer")
- parser.add_argument("--kernel_size",
- type=int,
- default=8,
- help="Kernel size for the (transposed) convolutions")
- parser.add_argument("--conv_stride",
- type=int,
- default=4,
- help="Stride for the (transposed) convolutions")
- parser.add_argument("--context",
- type=int,
- default=3,
- help="Context size for the decoder convolutions "
- "before the transposed convolutions")
- parser.add_argument("--rescale",
- type=float,
- default=0.1,
- help="Initial weight rescale reference")
- parser.add_argument("--no_resample", action="store_false",
- default=True, dest="resample",
- help="No Resampling of the input/output x2")
- parser.add_argument("--no_glu",
- action="store_false",
- default=True,
- dest="glu",
- help="Replace all GLUs by ReLUs")
- parser.add_argument("--no_rewrite",
- action="store_false",
- default=True,
- dest="rewrite",
- help="No 1x1 rewrite convolutions")
- parser.add_argument("--normalize", action="store_true")
- parser.add_argument("--no_norm_wav", action="store_false", dest='norm_wav', default=True)
-
- # Tasnet options
- parser.add_argument("--tasnet", action="store_true")
- parser.add_argument("--split_valid",
- action="store_true",
- help="Predict chunks by chunks for valid and test. Required for tasnet")
- parser.add_argument("--X", type=int, default=8)
-
- # Other options
- parser.add_argument("--show",
- action="store_true",
- help="Show model architecture, size and exit")
- parser.add_argument("--save_model", action="store_true",
- help="Skip traning, just save final model "
- "for the current checkpoint value.")
- parser.add_argument("--save_state",
- help="Skip training, just save state "
- "for the current checkpoint value. You should "
- "provide a model name as argument.")
-
- # Quantization options
- parser.add_argument("--q-min-size", type=float, default=1,
- help="Only quantize layers over this size (in MB)")
- parser.add_argument(
- "--qat", type=int, help="If provided, use QAT training with that many bits.")
-
- parser.add_argument("--diffq", type=float, default=0)
- parser.add_argument(
- "--ms-target", type=float, default=162,
- help="Model size target in MB, when using DiffQ. Best model will be kept "
- "only if it is smaller than this target.")
-
- return parser
-
-
-def get_name(parser, args):
- """
- Return the name of an experiment given the args. Some parameters are ignored,
- for instance --workers, as they do not impact the final result.
- """
- ignore_args = set([
- "checkpoints",
- "deterministic",
- "eval",
- "evals",
- "eval_cpu",
- "eval_workers",
- "logs",
- "master",
- "rank",
- "restart",
- "save",
- "save_model",
- "save_state",
- "show",
- "workers",
- "world_size",
- ])
- parts = []
- name_args = dict(args.__dict__)
- for name, value in name_args.items():
- if name in ignore_args:
- continue
- if value != parser.get_default(name):
- if isinstance(value, Path):
- parts.append(f"{name}={value.name}")
- else:
- parts.append(f"{name}={value}")
- if parts:
- name = " ".join(parts)
- else:
- name = "default"
- return name
diff --git a/demucs/pretrained.py b/demucs/pretrained.py
deleted file mode 100644
index 6aac5db100cc7a9084af96d2cd083f0c8fac473c..0000000000000000000000000000000000000000
--- a/demucs/pretrained.py
+++ /dev/null
@@ -1,107 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-# author: adefossez
-
-import logging
-
-from diffq import DiffQuantizer
-import torch.hub
-
-from .model import Demucs
-from .tasnet import ConvTasNet
-from .utils import set_state
-
-logger = logging.getLogger(__name__)
-ROOT = "https://dl.fbaipublicfiles.com/demucs/v3.0/"
-
-PRETRAINED_MODELS = {
- 'demucs': 'e07c671f',
- 'demucs48_hq': '28a1282c',
- 'demucs_extra': '3646af93',
- 'demucs_quantized': '07afea75',
- 'tasnet': 'beb46fac',
- 'tasnet_extra': 'df3777b2',
- 'demucs_unittest': '09ebc15f',
-}
-
-SOURCES = ["drums", "bass", "other", "vocals"]
-
-
-def get_url(name):
- sig = PRETRAINED_MODELS[name]
- return ROOT + name + "-" + sig[:8] + ".th"
-
-
-def is_pretrained(name):
- return name in PRETRAINED_MODELS
-
-
-def load_pretrained(name):
- if name == "demucs":
- return demucs(pretrained=True)
- elif name == "demucs48_hq":
- return demucs(pretrained=True, hq=True, channels=48)
- elif name == "demucs_extra":
- return demucs(pretrained=True, extra=True)
- elif name == "demucs_quantized":
- return demucs(pretrained=True, quantized=True)
- elif name == "demucs_unittest":
- return demucs_unittest(pretrained=True)
- elif name == "tasnet":
- return tasnet(pretrained=True)
- elif name == "tasnet_extra":
- return tasnet(pretrained=True, extra=True)
- else:
- raise ValueError(f"Invalid pretrained name {name}")
-
-
-def _load_state(name, model, quantizer=None):
- url = get_url(name)
- state = torch.hub.load_state_dict_from_url(url, map_location='cpu', check_hash=True)
- set_state(model, quantizer, state)
- if quantizer:
- quantizer.detach()
-
-
-def demucs_unittest(pretrained=True):
- model = Demucs(channels=4, sources=SOURCES)
- if pretrained:
- _load_state('demucs_unittest', model)
- return model
-
-
-def demucs(pretrained=True, extra=False, quantized=False, hq=False, channels=64):
- if not pretrained and (extra or quantized or hq):
- raise ValueError("if extra or quantized is True, pretrained must be True.")
- model = Demucs(sources=SOURCES, channels=channels)
- if pretrained:
- name = 'demucs'
- if channels != 64:
- name += str(channels)
- quantizer = None
- if sum([extra, quantized, hq]) > 1:
- raise ValueError("Only one of extra, quantized, hq, can be True.")
- if quantized:
- quantizer = DiffQuantizer(model, group_size=8, min_size=1)
- name += '_quantized'
- if extra:
- name += '_extra'
- if hq:
- name += '_hq'
- _load_state(name, model, quantizer)
- return model
-
-
-def tasnet(pretrained=True, extra=False):
- if not pretrained and extra:
- raise ValueError("if extra is True, pretrained must be True.")
- model = ConvTasNet(X=10, sources=SOURCES)
- if pretrained:
- name = 'tasnet'
- if extra:
- name = 'tasnet_extra'
- _load_state(name, model)
- return model
diff --git a/demucs/raw.py b/demucs/raw.py
deleted file mode 100644
index d4941ad2d7ed858f490db441f5b46b12bd61ad78..0000000000000000000000000000000000000000
--- a/demucs/raw.py
+++ /dev/null
@@ -1,173 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import argparse
-import os
-from collections import defaultdict, namedtuple
-from pathlib import Path
-
-import musdb
-import numpy as np
-import torch as th
-import tqdm
-from torch.utils.data import DataLoader
-
-from .audio import AudioFile
-
-ChunkInfo = namedtuple("ChunkInfo", ["file_index", "offset", "local_index"])
-
-
-class Rawset:
- """
- Dataset of raw, normalized, float32 audio files
- """
- def __init__(self, path, samples=None, stride=None, channels=2, streams=None):
- self.path = Path(path)
- self.channels = channels
- self.samples = samples
- if stride is None:
- stride = samples if samples is not None else 0
- self.stride = stride
- entries = defaultdict(list)
- for root, folders, files in os.walk(self.path, followlinks=True):
- folders.sort()
- files.sort()
- for file in files:
- if file.endswith(".raw"):
- path = Path(root) / file
- name, stream = path.stem.rsplit('.', 1)
- entries[(path.parent.relative_to(self.path), name)].append(int(stream))
-
- self._entries = list(entries.keys())
-
- sizes = []
- self._lengths = []
- ref_streams = sorted(entries[self._entries[0]])
- assert ref_streams == list(range(len(ref_streams)))
- if streams is None:
- self.streams = ref_streams
- else:
- self.streams = streams
- for entry in sorted(entries.keys()):
- streams = entries[entry]
- assert sorted(streams) == ref_streams
- file = self._path(*entry)
- length = file.stat().st_size // (4 * channels)
- if samples is None:
- sizes.append(1)
- else:
- if length < samples:
- self._entries.remove(entry)
- continue
- sizes.append((length - samples) // stride + 1)
- self._lengths.append(length)
- if not sizes:
- raise ValueError(f"Empty dataset {self.path}")
- self._cumulative_sizes = np.cumsum(sizes)
- self._sizes = sizes
-
- def __len__(self):
- return self._cumulative_sizes[-1]
-
- @property
- def total_length(self):
- return sum(self._lengths)
-
- def chunk_info(self, index):
- file_index = np.searchsorted(self._cumulative_sizes, index, side='right')
- if file_index == 0:
- local_index = index
- else:
- local_index = index - self._cumulative_sizes[file_index - 1]
- return ChunkInfo(offset=local_index * self.stride,
- file_index=file_index,
- local_index=local_index)
-
- def _path(self, folder, name, stream=0):
- return self.path / folder / (name + f'.{stream}.raw')
-
- def __getitem__(self, index):
- chunk = self.chunk_info(index)
- entry = self._entries[chunk.file_index]
-
- length = self.samples or self._lengths[chunk.file_index]
- streams = []
- to_read = length * self.channels * 4
- for stream_index, stream in enumerate(self.streams):
- offset = chunk.offset * 4 * self.channels
- file = open(self._path(*entry, stream=stream), 'rb')
- file.seek(offset)
- content = file.read(to_read)
- assert len(content) == to_read
- content = np.frombuffer(content, dtype=np.float32)
- content = content.copy() # make writable
- streams.append(th.from_numpy(content).view(length, self.channels).t())
- return th.stack(streams, dim=0)
-
- def name(self, index):
- chunk = self.chunk_info(index)
- folder, name = self._entries[chunk.file_index]
- return folder / name
-
-
-class MusDBSet:
- def __init__(self, mus, streams=slice(None), samplerate=44100, channels=2):
- self.mus = mus
- self.streams = streams
- self.samplerate = samplerate
- self.channels = channels
-
- def __len__(self):
- return len(self.mus.tracks)
-
- def __getitem__(self, index):
- track = self.mus.tracks[index]
- return (track.name, AudioFile(track.path).read(channels=self.channels,
- seek_time=0,
- streams=self.streams,
- samplerate=self.samplerate))
-
-
-def build_raw(mus, destination, normalize, workers, samplerate, channels):
- destination.mkdir(parents=True, exist_ok=True)
- loader = DataLoader(MusDBSet(mus, channels=channels, samplerate=samplerate),
- batch_size=1,
- num_workers=workers,
- collate_fn=lambda x: x[0])
- for name, streams in tqdm.tqdm(loader):
- if normalize:
- ref = streams[0].mean(dim=0) # use mono mixture as reference
- streams = (streams - ref.mean()) / ref.std()
- for index, stream in enumerate(streams):
- open(destination / (name + f'.{index}.raw'), "wb").write(stream.t().numpy().tobytes())
-
-
-def main():
- parser = argparse.ArgumentParser('rawset')
- parser.add_argument('--workers', type=int, default=10)
- parser.add_argument('--samplerate', type=int, default=44100)
- parser.add_argument('--channels', type=int, default=2)
- parser.add_argument('musdb', type=Path)
- parser.add_argument('destination', type=Path)
-
- args = parser.parse_args()
-
- build_raw(musdb.DB(root=args.musdb, subsets=["train"], split="train"),
- args.destination / "train",
- normalize=True,
- channels=args.channels,
- samplerate=args.samplerate,
- workers=args.workers)
- build_raw(musdb.DB(root=args.musdb, subsets=["train"], split="valid"),
- args.destination / "valid",
- normalize=True,
- samplerate=args.samplerate,
- channels=args.channels,
- workers=args.workers)
-
-
-if __name__ == "__main__":
- main()
diff --git a/demucs/repitch.py b/demucs/repitch.py
deleted file mode 100644
index 8846ab2d951a024c95067f66a113968500442828..0000000000000000000000000000000000000000
--- a/demucs/repitch.py
+++ /dev/null
@@ -1,96 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import io
-import random
-import subprocess as sp
-import tempfile
-
-import numpy as np
-import torch
-from scipy.io import wavfile
-
-
-def i16_pcm(wav):
- if wav.dtype == np.int16:
- return wav
- return (wav * 2**15).clamp_(-2**15, 2**15 - 1).short()
-
-
-def f32_pcm(wav):
- if wav.dtype == np.float:
- return wav
- return wav.float() / 2**15
-
-
-class RepitchedWrapper:
- """
- Wrap a dataset to apply online change of pitch / tempo.
- """
- def __init__(self, dataset, proba=0.2, max_pitch=2, max_tempo=12, tempo_std=5, vocals=[3]):
- self.dataset = dataset
- self.proba = proba
- self.max_pitch = max_pitch
- self.max_tempo = max_tempo
- self.tempo_std = tempo_std
- self.vocals = vocals
-
- def __len__(self):
- return len(self.dataset)
-
- def __getitem__(self, index):
- streams = self.dataset[index]
- in_length = streams.shape[-1]
- out_length = int((1 - 0.01 * self.max_tempo) * in_length)
-
- if random.random() < self.proba:
- delta_pitch = random.randint(-self.max_pitch, self.max_pitch)
- delta_tempo = random.gauss(0, self.tempo_std)
- delta_tempo = min(max(-self.max_tempo, delta_tempo), self.max_tempo)
- outs = []
- for idx, stream in enumerate(streams):
- stream = repitch(
- stream,
- delta_pitch,
- delta_tempo,
- voice=idx in self.vocals)
- outs.append(stream[:, :out_length])
- streams = torch.stack(outs)
- else:
- streams = streams[..., :out_length]
- return streams
-
-
-def repitch(wav, pitch, tempo, voice=False, quick=False, samplerate=44100):
- """
- tempo is a relative delta in percentage, so tempo=10 means tempo at 110%!
- pitch is in semi tones.
- Requires `soundstretch` to be installed, see
- https://www.surina.net/soundtouch/soundstretch.html
- """
- outfile = tempfile.NamedTemporaryFile(suffix=".wav")
- in_ = io.BytesIO()
- wavfile.write(in_, samplerate, i16_pcm(wav).t().numpy())
- command = [
- "soundstretch",
- "stdin",
- outfile.name,
- f"-pitch={pitch}",
- f"-tempo={tempo:.6f}",
- ]
- if quick:
- command += ["-quick"]
- if voice:
- command += ["-speech"]
- try:
- sp.run(command, capture_output=True, input=in_.getvalue(), check=True)
- except sp.CalledProcessError as error:
- raise RuntimeError(f"Could not change bpm because {error.stderr.decode('utf-8')}")
- sr, wav = wavfile.read(outfile.name)
- wav = wav.copy()
- wav = f32_pcm(torch.from_numpy(wav).t())
- assert sr == samplerate
- return wav
diff --git a/demucs/separate.py b/demucs/separate.py
deleted file mode 100644
index 3fc7af9e711978b3e21398aa6f1deb9ae87dd370..0000000000000000000000000000000000000000
--- a/demucs/separate.py
+++ /dev/null
@@ -1,185 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import argparse
-import sys
-from pathlib import Path
-import subprocess
-
-import julius
-import torch as th
-import torchaudio as ta
-
-from .audio import AudioFile, convert_audio_channels
-from .pretrained import is_pretrained, load_pretrained
-from .utils import apply_model, load_model
-
-
-def load_track(track, device, audio_channels, samplerate):
- errors = {}
- wav = None
-
- try:
- wav = AudioFile(track).read(
- streams=0,
- samplerate=samplerate,
- channels=audio_channels).to(device)
- except FileNotFoundError:
- errors['ffmpeg'] = 'Ffmpeg is not installed.'
- except subprocess.CalledProcessError:
- errors['ffmpeg'] = 'FFmpeg could not read the file.'
-
- if wav is None:
- try:
- wav, sr = ta.load(str(track))
- except RuntimeError as err:
- errors['torchaudio'] = err.args[0]
- else:
- wav = convert_audio_channels(wav, audio_channels)
- wav = wav.to(device)
- wav = julius.resample_frac(wav, sr, samplerate)
-
- if wav is None:
- print(f"Could not load file {track}. "
- "Maybe it is not a supported file format? ")
- for backend, error in errors.items():
- print(f"When trying to load using {backend}, got the following error: {error}")
- sys.exit(1)
- return wav
-
-
-def encode_mp3(wav, path, bitrate=320, samplerate=44100, channels=2, verbose=False):
- try:
- import lameenc
- except ImportError:
- print("Failed to call lame encoder. Maybe it is not installed? "
- "On windows, run `python.exe -m pip install -U lameenc`, "
- "on OSX/Linux, run `python3 -m pip install -U lameenc`, "
- "then try again.", file=sys.stderr)
- sys.exit(1)
- encoder = lameenc.Encoder()
- encoder.set_bit_rate(bitrate)
- encoder.set_in_sample_rate(samplerate)
- encoder.set_channels(channels)
- encoder.set_quality(2) # 2-highest, 7-fastest
- if not verbose:
- encoder.silence()
- wav = wav.transpose(0, 1).numpy()
- mp3_data = encoder.encode(wav.tobytes())
- mp3_data += encoder.flush()
- with open(path, "wb") as f:
- f.write(mp3_data)
-
-
-def main():
- parser = argparse.ArgumentParser("demucs.separate",
- description="Separate the sources for the given tracks")
- parser.add_argument("tracks", nargs='+', type=Path, default=[], help='Path to tracks')
- parser.add_argument("-n",
- "--name",
- default="demucs_quantized",
- help="Model name. See README.md for the list of pretrained models. "
- "Default is demucs_quantized.")
- parser.add_argument("-v", "--verbose", action="store_true")
- parser.add_argument("-o",
- "--out",
- type=Path,
- default=Path("separated"),
- help="Folder where to put extracted tracks. A subfolder "
- "with the model name will be created.")
- parser.add_argument("--models",
- type=Path,
- default=Path("models"),
- help="Path to trained models. "
- "Also used to store downloaded pretrained models")
- parser.add_argument("-d",
- "--device",
- default="cuda" if th.cuda.is_available() else "cpu",
- help="Device to use, default is cuda if available else cpu")
- parser.add_argument("--shifts",
- default=0,
- type=int,
- help="Number of random shifts for equivariant stabilization."
- "Increase separation time but improves quality for Demucs. 10 was used "
- "in the original paper.")
- parser.add_argument("--overlap",
- default=0.25,
- type=float,
- help="Overlap between the splits.")
- parser.add_argument("--no-split",
- action="store_false",
- dest="split",
- default=True,
- help="Doesn't split audio in chunks. This can use large amounts of memory.")
- parser.add_argument("--float32",
- action="store_true",
- help="Convert the output wavefile to use pcm f32 format instead of s16. "
- "This should not make a difference if you just plan on listening to the "
- "audio but might be needed to compute exactly metrics like SDR etc.")
- parser.add_argument("--int16",
- action="store_false",
- dest="float32",
- help="Opposite of --float32, here for compatibility.")
- parser.add_argument("--mp3", action="store_true",
- help="Convert the output wavs to mp3.")
- parser.add_argument("--mp3-bitrate",
- default=320,
- type=int,
- help="Bitrate of converted mp3.")
-
- args = parser.parse_args()
- name = args.name + ".th"
- model_path = args.models / name
- if model_path.is_file():
- model = load_model(model_path)
- else:
- if is_pretrained(args.name):
- model = load_pretrained(args.name)
- else:
- print(f"No pre-trained model {args.name}", file=sys.stderr)
- sys.exit(1)
- model.to(args.device)
-
- out = args.out / args.name
- out.mkdir(parents=True, exist_ok=True)
- print(f"Separated tracks will be stored in {out.resolve()}")
- for track in args.tracks:
- if not track.exists():
- print(
- f"File {track} does not exist. If the path contains spaces, "
- "please try again after surrounding the entire path with quotes \"\".",
- file=sys.stderr)
- continue
- print(f"Separating track {track}")
- wav = load_track(track, args.device, model.audio_channels, model.samplerate)
-
- ref = wav.mean(0)
- wav = (wav - ref.mean()) / ref.std()
- sources = apply_model(model, wav, shifts=args.shifts, split=args.split,
- overlap=args.overlap, progress=True)
- sources = sources * ref.std() + ref.mean()
-
- track_folder = out / track.name.rsplit(".", 1)[0]
- track_folder.mkdir(exist_ok=True)
- for source, name in zip(sources, model.sources):
- source = source / max(1.01 * source.abs().max(), 1)
- if args.mp3 or not args.float32:
- source = (source * 2**15).clamp_(-2**15, 2**15 - 1).short()
- source = source.cpu()
- stem = str(track_folder / name)
- if args.mp3:
- encode_mp3(source, stem + ".mp3",
- bitrate=args.mp3_bitrate,
- samplerate=model.samplerate,
- channels=model.audio_channels,
- verbose=args.verbose)
- else:
- wavname = str(track_folder / f"{name}.wav")
- ta.save(wavname, source, sample_rate=model.samplerate)
-
-
-if __name__ == "__main__":
- main()
diff --git a/demucs/tasnet.py b/demucs/tasnet.py
deleted file mode 100644
index ecc1257925ea8f4fbe389ddd6d73ce9fdf45f6d4..0000000000000000000000000000000000000000
--- a/demucs/tasnet.py
+++ /dev/null
@@ -1,452 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-#
-# Created on 2018/12
-# Author: Kaituo XU
-# Modified on 2019/11 by Alexandre Defossez, added support for multiple output channels
-# Here is the original license:
-# The MIT License (MIT)
-#
-# Copyright (c) 2018 Kaituo XU
-#
-# Permission is hereby granted, free of charge, to any person obtaining a copy
-# of this software and associated documentation files (the "Software"), to deal
-# in the Software without restriction, including without limitation the rights
-# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-# copies of the Software, and to permit persons to whom the Software is
-# furnished to do so, subject to the following conditions:
-#
-# The above copyright notice and this permission notice shall be included in all
-# copies or substantial portions of the Software.
-#
-# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-# SOFTWARE.
-
-import math
-
-import torch
-import torch.nn as nn
-import torch.nn.functional as F
-
-from .utils import capture_init
-
-EPS = 1e-8
-
-
-def overlap_and_add(signal, frame_step):
- outer_dimensions = signal.size()[:-2]
- frames, frame_length = signal.size()[-2:]
-
- subframe_length = math.gcd(frame_length, frame_step) # gcd=Greatest Common Divisor
- subframe_step = frame_step // subframe_length
- subframes_per_frame = frame_length // subframe_length
- output_size = frame_step * (frames - 1) + frame_length
- output_subframes = output_size // subframe_length
-
- subframe_signal = signal.view(*outer_dimensions, -1, subframe_length)
-
- frame = torch.arange(0, output_subframes,
- device=signal.device).unfold(0, subframes_per_frame, subframe_step)
- frame = frame.long() # signal may in GPU or CPU
- frame = frame.contiguous().view(-1)
-
- result = signal.new_zeros(*outer_dimensions, output_subframes, subframe_length)
- result.index_add_(-2, frame, subframe_signal)
- result = result.view(*outer_dimensions, -1)
- return result
-
-
-class ConvTasNet(nn.Module):
- @capture_init
- def __init__(self,
- sources,
- N=256,
- L=20,
- B=256,
- H=512,
- P=3,
- X=8,
- R=4,
- audio_channels=2,
- norm_type="gLN",
- causal=False,
- mask_nonlinear='relu',
- samplerate=44100,
- segment_length=44100 * 2 * 4):
- """
- Args:
- sources: list of sources
- N: Number of filters in autoencoder
- L: Length of the filters (in samples)
- B: Number of channels in bottleneck 1 × 1-conv block
- H: Number of channels in convolutional blocks
- P: Kernel size in convolutional blocks
- X: Number of convolutional blocks in each repeat
- R: Number of repeats
- norm_type: BN, gLN, cLN
- causal: causal or non-causal
- mask_nonlinear: use which non-linear function to generate mask
- """
- super(ConvTasNet, self).__init__()
- # Hyper-parameter
- self.sources = sources
- self.C = len(sources)
- self.N, self.L, self.B, self.H, self.P, self.X, self.R = N, L, B, H, P, X, R
- self.norm_type = norm_type
- self.causal = causal
- self.mask_nonlinear = mask_nonlinear
- self.audio_channels = audio_channels
- self.samplerate = samplerate
- self.segment_length = segment_length
- # Components
- self.encoder = Encoder(L, N, audio_channels)
- self.separator = TemporalConvNet(
- N, B, H, P, X, R, self.C, norm_type, causal, mask_nonlinear)
- self.decoder = Decoder(N, L, audio_channels)
- # init
- for p in self.parameters():
- if p.dim() > 1:
- nn.init.xavier_normal_(p)
-
- def valid_length(self, length):
- return length
-
- def forward(self, mixture):
- """
- Args:
- mixture: [M, T], M is batch size, T is #samples
- Returns:
- est_source: [M, C, T]
- """
- mixture_w = self.encoder(mixture)
- est_mask = self.separator(mixture_w)
- est_source = self.decoder(mixture_w, est_mask)
-
- # T changed after conv1d in encoder, fix it here
- T_origin = mixture.size(-1)
- T_conv = est_source.size(-1)
- est_source = F.pad(est_source, (0, T_origin - T_conv))
- return est_source
-
-
-class Encoder(nn.Module):
- """Estimation of the nonnegative mixture weight by a 1-D conv layer.
- """
- def __init__(self, L, N, audio_channels):
- super(Encoder, self).__init__()
- # Hyper-parameter
- self.L, self.N = L, N
- # Components
- # 50% overlap
- self.conv1d_U = nn.Conv1d(audio_channels, N, kernel_size=L, stride=L // 2, bias=False)
-
- def forward(self, mixture):
- """
- Args:
- mixture: [M, T], M is batch size, T is #samples
- Returns:
- mixture_w: [M, N, K], where K = (T-L)/(L/2)+1 = 2T/L-1
- """
- mixture_w = F.relu(self.conv1d_U(mixture)) # [M, N, K]
- return mixture_w
-
-
-class Decoder(nn.Module):
- def __init__(self, N, L, audio_channels):
- super(Decoder, self).__init__()
- # Hyper-parameter
- self.N, self.L = N, L
- self.audio_channels = audio_channels
- # Components
- self.basis_signals = nn.Linear(N, audio_channels * L, bias=False)
-
- def forward(self, mixture_w, est_mask):
- """
- Args:
- mixture_w: [M, N, K]
- est_mask: [M, C, N, K]
- Returns:
- est_source: [M, C, T]
- """
- # D = W * M
- source_w = torch.unsqueeze(mixture_w, 1) * est_mask # [M, C, N, K]
- source_w = torch.transpose(source_w, 2, 3) # [M, C, K, N]
- # S = DV
- est_source = self.basis_signals(source_w) # [M, C, K, ac * L]
- m, c, k, _ = est_source.size()
- est_source = est_source.view(m, c, k, self.audio_channels, -1).transpose(2, 3).contiguous()
- est_source = overlap_and_add(est_source, self.L // 2) # M x C x ac x T
- return est_source
-
-
-class TemporalConvNet(nn.Module):
- def __init__(self, N, B, H, P, X, R, C, norm_type="gLN", causal=False, mask_nonlinear='relu'):
- """
- Args:
- N: Number of filters in autoencoder
- B: Number of channels in bottleneck 1 × 1-conv block
- H: Number of channels in convolutional blocks
- P: Kernel size in convolutional blocks
- X: Number of convolutional blocks in each repeat
- R: Number of repeats
- C: Number of speakers
- norm_type: BN, gLN, cLN
- causal: causal or non-causal
- mask_nonlinear: use which non-linear function to generate mask
- """
- super(TemporalConvNet, self).__init__()
- # Hyper-parameter
- self.C = C
- self.mask_nonlinear = mask_nonlinear
- # Components
- # [M, N, K] -> [M, N, K]
- layer_norm = ChannelwiseLayerNorm(N)
- # [M, N, K] -> [M, B, K]
- bottleneck_conv1x1 = nn.Conv1d(N, B, 1, bias=False)
- # [M, B, K] -> [M, B, K]
- repeats = []
- for r in range(R):
- blocks = []
- for x in range(X):
- dilation = 2**x
- padding = (P - 1) * dilation if causal else (P - 1) * dilation // 2
- blocks += [
- TemporalBlock(B,
- H,
- P,
- stride=1,
- padding=padding,
- dilation=dilation,
- norm_type=norm_type,
- causal=causal)
- ]
- repeats += [nn.Sequential(*blocks)]
- temporal_conv_net = nn.Sequential(*repeats)
- # [M, B, K] -> [M, C*N, K]
- mask_conv1x1 = nn.Conv1d(B, C * N, 1, bias=False)
- # Put together
- self.network = nn.Sequential(layer_norm, bottleneck_conv1x1, temporal_conv_net,
- mask_conv1x1)
-
- def forward(self, mixture_w):
- """
- Keep this API same with TasNet
- Args:
- mixture_w: [M, N, K], M is batch size
- returns:
- est_mask: [M, C, N, K]
- """
- M, N, K = mixture_w.size()
- score = self.network(mixture_w) # [M, N, K] -> [M, C*N, K]
- score = score.view(M, self.C, N, K) # [M, C*N, K] -> [M, C, N, K]
- if self.mask_nonlinear == 'softmax':
- est_mask = F.softmax(score, dim=1)
- elif self.mask_nonlinear == 'relu':
- est_mask = F.relu(score)
- else:
- raise ValueError("Unsupported mask non-linear function")
- return est_mask
-
-
-class TemporalBlock(nn.Module):
- def __init__(self,
- in_channels,
- out_channels,
- kernel_size,
- stride,
- padding,
- dilation,
- norm_type="gLN",
- causal=False):
- super(TemporalBlock, self).__init__()
- # [M, B, K] -> [M, H, K]
- conv1x1 = nn.Conv1d(in_channels, out_channels, 1, bias=False)
- prelu = nn.PReLU()
- norm = chose_norm(norm_type, out_channels)
- # [M, H, K] -> [M, B, K]
- dsconv = DepthwiseSeparableConv(out_channels, in_channels, kernel_size, stride, padding,
- dilation, norm_type, causal)
- # Put together
- self.net = nn.Sequential(conv1x1, prelu, norm, dsconv)
-
- def forward(self, x):
- """
- Args:
- x: [M, B, K]
- Returns:
- [M, B, K]
- """
- residual = x
- out = self.net(x)
- # TODO: when P = 3 here works fine, but when P = 2 maybe need to pad?
- return out + residual # look like w/o F.relu is better than w/ F.relu
- # return F.relu(out + residual)
-
-
-class DepthwiseSeparableConv(nn.Module):
- def __init__(self,
- in_channels,
- out_channels,
- kernel_size,
- stride,
- padding,
- dilation,
- norm_type="gLN",
- causal=False):
- super(DepthwiseSeparableConv, self).__init__()
- # Use `groups` option to implement depthwise convolution
- # [M, H, K] -> [M, H, K]
- depthwise_conv = nn.Conv1d(in_channels,
- in_channels,
- kernel_size,
- stride=stride,
- padding=padding,
- dilation=dilation,
- groups=in_channels,
- bias=False)
- if causal:
- chomp = Chomp1d(padding)
- prelu = nn.PReLU()
- norm = chose_norm(norm_type, in_channels)
- # [M, H, K] -> [M, B, K]
- pointwise_conv = nn.Conv1d(in_channels, out_channels, 1, bias=False)
- # Put together
- if causal:
- self.net = nn.Sequential(depthwise_conv, chomp, prelu, norm, pointwise_conv)
- else:
- self.net = nn.Sequential(depthwise_conv, prelu, norm, pointwise_conv)
-
- def forward(self, x):
- """
- Args:
- x: [M, H, K]
- Returns:
- result: [M, B, K]
- """
- return self.net(x)
-
-
-class Chomp1d(nn.Module):
- """To ensure the output length is the same as the input.
- """
- def __init__(self, chomp_size):
- super(Chomp1d, self).__init__()
- self.chomp_size = chomp_size
-
- def forward(self, x):
- """
- Args:
- x: [M, H, Kpad]
- Returns:
- [M, H, K]
- """
- return x[:, :, :-self.chomp_size].contiguous()
-
-
-def chose_norm(norm_type, channel_size):
- """The input of normlization will be (M, C, K), where M is batch size,
- C is channel size and K is sequence length.
- """
- if norm_type == "gLN":
- return GlobalLayerNorm(channel_size)
- elif norm_type == "cLN":
- return ChannelwiseLayerNorm(channel_size)
- elif norm_type == "id":
- return nn.Identity()
- else: # norm_type == "BN":
- # Given input (M, C, K), nn.BatchNorm1d(C) will accumulate statics
- # along M and K, so this BN usage is right.
- return nn.BatchNorm1d(channel_size)
-
-
-# TODO: Use nn.LayerNorm to impl cLN to speed up
-class ChannelwiseLayerNorm(nn.Module):
- """Channel-wise Layer Normalization (cLN)"""
- def __init__(self, channel_size):
- super(ChannelwiseLayerNorm, self).__init__()
- self.gamma = nn.Parameter(torch.Tensor(1, channel_size, 1)) # [1, N, 1]
- self.beta = nn.Parameter(torch.Tensor(1, channel_size, 1)) # [1, N, 1]
- self.reset_parameters()
-
- def reset_parameters(self):
- self.gamma.data.fill_(1)
- self.beta.data.zero_()
-
- def forward(self, y):
- """
- Args:
- y: [M, N, K], M is batch size, N is channel size, K is length
- Returns:
- cLN_y: [M, N, K]
- """
- mean = torch.mean(y, dim=1, keepdim=True) # [M, 1, K]
- var = torch.var(y, dim=1, keepdim=True, unbiased=False) # [M, 1, K]
- cLN_y = self.gamma * (y - mean) / torch.pow(var + EPS, 0.5) + self.beta
- return cLN_y
-
-
-class GlobalLayerNorm(nn.Module):
- """Global Layer Normalization (gLN)"""
- def __init__(self, channel_size):
- super(GlobalLayerNorm, self).__init__()
- self.gamma = nn.Parameter(torch.Tensor(1, channel_size, 1)) # [1, N, 1]
- self.beta = nn.Parameter(torch.Tensor(1, channel_size, 1)) # [1, N, 1]
- self.reset_parameters()
-
- def reset_parameters(self):
- self.gamma.data.fill_(1)
- self.beta.data.zero_()
-
- def forward(self, y):
- """
- Args:
- y: [M, N, K], M is batch size, N is channel size, K is length
- Returns:
- gLN_y: [M, N, K]
- """
- # TODO: in torch 1.0, torch.mean() support dim list
- mean = y.mean(dim=1, keepdim=True).mean(dim=2, keepdim=True) # [M, 1, 1]
- var = (torch.pow(y - mean, 2)).mean(dim=1, keepdim=True).mean(dim=2, keepdim=True)
- gLN_y = self.gamma * (y - mean) / torch.pow(var + EPS, 0.5) + self.beta
- return gLN_y
-
-
-if __name__ == "__main__":
- torch.manual_seed(123)
- M, N, L, T = 2, 3, 4, 12
- K = 2 * T // L - 1
- B, H, P, X, R, C, norm_type, causal = 2, 3, 3, 3, 2, 2, "gLN", False
- mixture = torch.randint(3, (M, T))
- # test Encoder
- encoder = Encoder(L, N)
- encoder.conv1d_U.weight.data = torch.randint(2, encoder.conv1d_U.weight.size())
- mixture_w = encoder(mixture)
- print('mixture', mixture)
- print('U', encoder.conv1d_U.weight)
- print('mixture_w', mixture_w)
- print('mixture_w size', mixture_w.size())
-
- # test TemporalConvNet
- separator = TemporalConvNet(N, B, H, P, X, R, C, norm_type=norm_type, causal=causal)
- est_mask = separator(mixture_w)
- print('est_mask', est_mask)
-
- # test Decoder
- decoder = Decoder(N, L)
- est_mask = torch.randint(2, (B, K, C, N))
- est_source = decoder(mixture_w, est_mask)
- print('est_source', est_source)
-
- # test Conv-TasNet
- conv_tasnet = ConvTasNet(N, L, B, H, P, X, R, C, norm_type=norm_type)
- est_source = conv_tasnet(mixture)
- print('est_source', est_source)
- print('est_source size', est_source.size())
diff --git a/demucs/test.py b/demucs/test.py
deleted file mode 100644
index 4140914ddbff3543b4056ca0cb1b5e887434a40a..0000000000000000000000000000000000000000
--- a/demucs/test.py
+++ /dev/null
@@ -1,109 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import gzip
-import sys
-from concurrent import futures
-
-import musdb
-import museval
-import torch as th
-import tqdm
-from scipy.io import wavfile
-from torch import distributed
-
-from .audio import convert_audio
-from .utils import apply_model
-
-
-def evaluate(model,
- musdb_path,
- eval_folder,
- workers=2,
- device="cpu",
- rank=0,
- save=False,
- shifts=0,
- split=False,
- overlap=0.25,
- is_wav=False,
- world_size=1):
- """
- Evaluate model using museval. Run the model
- on a single GPU, the bottleneck being the call to museval.
- """
-
- output_dir = eval_folder / "results"
- output_dir.mkdir(exist_ok=True, parents=True)
- json_folder = eval_folder / "results/test"
- json_folder.mkdir(exist_ok=True, parents=True)
-
- # we load tracks from the original musdb set
- test_set = musdb.DB(musdb_path, subsets=["test"], is_wav=is_wav)
- src_rate = 44100 # hardcoded for now...
-
- for p in model.parameters():
- p.requires_grad = False
- p.grad = None
-
- pendings = []
- with futures.ProcessPoolExecutor(workers or 1) as pool:
- for index in tqdm.tqdm(range(rank, len(test_set), world_size), file=sys.stdout):
- track = test_set.tracks[index]
-
- out = json_folder / f"{track.name}.json.gz"
- if out.exists():
- continue
-
- mix = th.from_numpy(track.audio).t().float()
- ref = mix.mean(dim=0) # mono mixture
- mix = (mix - ref.mean()) / ref.std()
- mix = convert_audio(mix, src_rate, model.samplerate, model.audio_channels)
- estimates = apply_model(model, mix.to(device),
- shifts=shifts, split=split, overlap=overlap)
- estimates = estimates * ref.std() + ref.mean()
-
- estimates = estimates.transpose(1, 2)
- references = th.stack(
- [th.from_numpy(track.targets[name].audio).t() for name in model.sources])
- references = convert_audio(references, src_rate,
- model.samplerate, model.audio_channels)
- references = references.transpose(1, 2).numpy()
- estimates = estimates.cpu().numpy()
- win = int(1. * model.samplerate)
- hop = int(1. * model.samplerate)
- if save:
- folder = eval_folder / "wav/test" / track.name
- folder.mkdir(exist_ok=True, parents=True)
- for name, estimate in zip(model.sources, estimates):
- wavfile.write(str(folder / (name + ".wav")), 44100, estimate)
-
- if workers:
- pendings.append((track.name, pool.submit(
- museval.evaluate, references, estimates, win=win, hop=hop)))
- else:
- pendings.append((track.name, museval.evaluate(
- references, estimates, win=win, hop=hop)))
- del references, mix, estimates, track
-
- for track_name, pending in tqdm.tqdm(pendings, file=sys.stdout):
- if workers:
- pending = pending.result()
- sdr, isr, sir, sar = pending
- track_store = museval.TrackStore(win=44100, hop=44100, track_name=track_name)
- for idx, target in enumerate(model.sources):
- values = {
- "SDR": sdr[idx].tolist(),
- "SIR": sir[idx].tolist(),
- "ISR": isr[idx].tolist(),
- "SAR": sar[idx].tolist()
- }
-
- track_store.add_target(target_name=target, values=values)
- json_path = json_folder / f"{track_name}.json.gz"
- gzip.open(json_path, "w").write(track_store.json.encode('utf-8'))
- if world_size > 1:
- distributed.barrier()
diff --git a/demucs/train.py b/demucs/train.py
deleted file mode 100644
index 6bd221279dc986a6df1a8d7b4d4444bb822a1cb3..0000000000000000000000000000000000000000
--- a/demucs/train.py
+++ /dev/null
@@ -1,127 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import sys
-
-import tqdm
-from torch.utils.data import DataLoader
-from torch.utils.data.distributed import DistributedSampler
-
-from .utils import apply_model, average_metric, center_trim
-
-
-def train_model(epoch,
- dataset,
- model,
- criterion,
- optimizer,
- augment,
- quantizer=None,
- diffq=0,
- repeat=1,
- device="cpu",
- seed=None,
- workers=4,
- world_size=1,
- batch_size=16):
-
- if world_size > 1:
- sampler = DistributedSampler(dataset)
- sampler_epoch = epoch * repeat
- if seed is not None:
- sampler_epoch += seed * 1000
- sampler.set_epoch(sampler_epoch)
- batch_size //= world_size
- loader = DataLoader(dataset, batch_size=batch_size, sampler=sampler, num_workers=workers)
- else:
- loader = DataLoader(dataset, batch_size=batch_size, num_workers=workers, shuffle=True)
- current_loss = 0
- model_size = 0
- for repetition in range(repeat):
- tq = tqdm.tqdm(loader,
- ncols=120,
- desc=f"[{epoch:03d}] train ({repetition + 1}/{repeat})",
- leave=False,
- file=sys.stdout,
- unit=" batch")
- total_loss = 0
- for idx, sources in enumerate(tq):
- if len(sources) < batch_size:
- # skip uncomplete batch for augment.Remix to work properly
- continue
- sources = sources.to(device)
- sources = augment(sources)
- mix = sources.sum(dim=1)
-
- estimates = model(mix)
- sources = center_trim(sources, estimates)
- loss = criterion(estimates, sources)
- model_size = 0
- if quantizer is not None:
- model_size = quantizer.model_size()
-
- train_loss = loss + diffq * model_size
- train_loss.backward()
- grad_norm = 0
- for p in model.parameters():
- if p.grad is not None:
- grad_norm += p.grad.data.norm()**2
- grad_norm = grad_norm**0.5
- optimizer.step()
- optimizer.zero_grad()
-
- if quantizer is not None:
- model_size = model_size.item()
-
- total_loss += loss.item()
- current_loss = total_loss / (1 + idx)
- tq.set_postfix(loss=f"{current_loss:.4f}", ms=f"{model_size:.2f}",
- grad=f"{grad_norm:.5f}")
-
- # free some space before next round
- del sources, mix, estimates, loss, train_loss
-
- if world_size > 1:
- sampler.epoch += 1
-
- if world_size > 1:
- current_loss = average_metric(current_loss)
- return current_loss, model_size
-
-
-def validate_model(epoch,
- dataset,
- model,
- criterion,
- device="cpu",
- rank=0,
- world_size=1,
- shifts=0,
- overlap=0.25,
- split=False):
- indexes = range(rank, len(dataset), world_size)
- tq = tqdm.tqdm(indexes,
- ncols=120,
- desc=f"[{epoch:03d}] valid",
- leave=False,
- file=sys.stdout,
- unit=" track")
- current_loss = 0
- for index in tq:
- streams = dataset[index]
- # first five minutes to avoid OOM on --upsample models
- streams = streams[..., :15_000_000]
- streams = streams.to(device)
- sources = streams[1:]
- mix = streams[0]
- estimates = apply_model(model, mix, shifts=shifts, split=split, overlap=overlap)
- loss = criterion(estimates, sources)
- current_loss += loss.item() / len(indexes)
- del estimates, streams, sources
-
- if world_size > 1:
- current_loss = average_metric(current_loss, len(indexes))
- return current_loss
diff --git a/demucs/utils.py b/demucs/utils.py
deleted file mode 100644
index 4364184059b1afe3c8379c77793a8e76dccf9699..0000000000000000000000000000000000000000
--- a/demucs/utils.py
+++ /dev/null
@@ -1,323 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import errno
-import functools
-import hashlib
-import inspect
-import io
-import os
-import random
-import socket
-import tempfile
-import warnings
-import zlib
-from contextlib import contextmanager
-
-from diffq import UniformQuantizer, DiffQuantizer
-import torch as th
-import tqdm
-from torch import distributed
-from torch.nn import functional as F
-
-
-def center_trim(tensor, reference):
- """
- Center trim `tensor` with respect to `reference`, along the last dimension.
- `reference` can also be a number, representing the length to trim to.
- If the size difference != 0 mod 2, the extra sample is removed on the right side.
- """
- if hasattr(reference, "size"):
- reference = reference.size(-1)
- delta = tensor.size(-1) - reference
- if delta < 0:
- raise ValueError("tensor must be larger than reference. " f"Delta is {delta}.")
- if delta:
- tensor = tensor[..., delta // 2:-(delta - delta // 2)]
- return tensor
-
-
-def average_metric(metric, count=1.):
- """
- Average `metric` which should be a float across all hosts. `count` should be
- the weight for this particular host (i.e. number of examples).
- """
- metric = th.tensor([count, count * metric], dtype=th.float32, device='cuda')
- distributed.all_reduce(metric, op=distributed.ReduceOp.SUM)
- return metric[1].item() / metric[0].item()
-
-
-def free_port(host='', low=20000, high=40000):
- """
- Return a port number that is most likely free.
- This could suffer from a race condition although
- it should be quite rare.
- """
- sock = socket.socket()
- while True:
- port = random.randint(low, high)
- try:
- sock.bind((host, port))
- except OSError as error:
- if error.errno == errno.EADDRINUSE:
- continue
- raise
- return port
-
-
-def sizeof_fmt(num, suffix='B'):
- """
- Given `num` bytes, return human readable size.
- Taken from https://stackoverflow.com/a/1094933
- """
- for unit in ['', 'Ki', 'Mi', 'Gi', 'Ti', 'Pi', 'Ei', 'Zi']:
- if abs(num) < 1024.0:
- return "%3.1f%s%s" % (num, unit, suffix)
- num /= 1024.0
- return "%.1f%s%s" % (num, 'Yi', suffix)
-
-
-def human_seconds(seconds, display='.2f'):
- """
- Given `seconds` seconds, return human readable duration.
- """
- value = seconds * 1e6
- ratios = [1e3, 1e3, 60, 60, 24]
- names = ['us', 'ms', 's', 'min', 'hrs', 'days']
- last = names.pop(0)
- for name, ratio in zip(names, ratios):
- if value / ratio < 0.3:
- break
- value /= ratio
- last = name
- return f"{format(value, display)} {last}"
-
-
-class TensorChunk:
- def __init__(self, tensor, offset=0, length=None):
- total_length = tensor.shape[-1]
- assert offset >= 0
- assert offset < total_length
-
- if length is None:
- length = total_length - offset
- else:
- length = min(total_length - offset, length)
-
- self.tensor = tensor
- self.offset = offset
- self.length = length
- self.device = tensor.device
-
- @property
- def shape(self):
- shape = list(self.tensor.shape)
- shape[-1] = self.length
- return shape
-
- def padded(self, target_length):
- delta = target_length - self.length
- total_length = self.tensor.shape[-1]
- assert delta >= 0
-
- start = self.offset - delta // 2
- end = start + target_length
-
- correct_start = max(0, start)
- correct_end = min(total_length, end)
-
- pad_left = correct_start - start
- pad_right = end - correct_end
-
- out = F.pad(self.tensor[..., correct_start:correct_end], (pad_left, pad_right))
- assert out.shape[-1] == target_length
- return out
-
-
-def tensor_chunk(tensor_or_chunk):
- if isinstance(tensor_or_chunk, TensorChunk):
- return tensor_or_chunk
- else:
- assert isinstance(tensor_or_chunk, th.Tensor)
- return TensorChunk(tensor_or_chunk)
-
-
-def apply_model(model, mix, shifts=None, split=False,
- overlap=0.25, transition_power=1., progress=False):
- """
- Apply model to a given mixture.
-
- Args:
- shifts (int): if > 0, will shift in time `mix` by a random amount between 0 and 0.5 sec
- and apply the oppositve shift to the output. This is repeated `shifts` time and
- all predictions are averaged. This effectively makes the model time equivariant
- and improves SDR by up to 0.2 points.
- split (bool): if True, the input will be broken down in 8 seconds extracts
- and predictions will be performed individually on each and concatenated.
- Useful for model with large memory footprint like Tasnet.
- progress (bool): if True, show a progress bar (requires split=True)
- """
- assert transition_power >= 1, "transition_power < 1 leads to weird behavior."
- device = mix.device
- channels, length = mix.shape
- if split:
- out = th.zeros(len(model.sources), channels, length, device=device)
- sum_weight = th.zeros(length, device=device)
- segment = model.segment_length
- stride = int((1 - overlap) * segment)
- offsets = range(0, length, stride)
- scale = stride / model.samplerate
- if progress:
- offsets = tqdm.tqdm(offsets, unit_scale=scale, ncols=120, unit='seconds')
- # We start from a triangle shaped weight, with maximal weight in the middle
- # of the segment. Then we normalize and take to the power `transition_power`.
- # Large values of transition power will lead to sharper transitions.
- weight = th.cat([th.arange(1, segment // 2 + 1),
- th.arange(segment - segment // 2, 0, -1)]).to(device)
- assert len(weight) == segment
- # If the overlap < 50%, this will translate to linear transition when
- # transition_power is 1.
- weight = (weight / weight.max())**transition_power
- for offset in offsets:
- chunk = TensorChunk(mix, offset, segment)
- chunk_out = apply_model(model, chunk, shifts=shifts)
- chunk_length = chunk_out.shape[-1]
- out[..., offset:offset + segment] += weight[:chunk_length] * chunk_out
- sum_weight[offset:offset + segment] += weight[:chunk_length]
- offset += segment
- assert sum_weight.min() > 0
- out /= sum_weight
- return out
- elif shifts:
- max_shift = int(0.5 * model.samplerate)
- mix = tensor_chunk(mix)
- padded_mix = mix.padded(length + 2 * max_shift)
- out = 0
- for _ in range(shifts):
- offset = random.randint(0, max_shift)
- shifted = TensorChunk(padded_mix, offset, length + max_shift - offset)
- shifted_out = apply_model(model, shifted)
- out += shifted_out[..., max_shift - offset:]
- out /= shifts
- return out
- else:
- valid_length = model.valid_length(length)
- mix = tensor_chunk(mix)
- padded_mix = mix.padded(valid_length)
- with th.no_grad():
- out = model(padded_mix.unsqueeze(0))[0]
- return center_trim(out, length)
-
-
-@contextmanager
-def temp_filenames(count, delete=True):
- names = []
- try:
- for _ in range(count):
- names.append(tempfile.NamedTemporaryFile(delete=False).name)
- yield names
- finally:
- if delete:
- for name in names:
- os.unlink(name)
-
-
-def get_quantizer(model, args, optimizer=None):
- quantizer = None
- if args.diffq:
- quantizer = DiffQuantizer(
- model, min_size=args.q_min_size, group_size=8)
- if optimizer is not None:
- quantizer.setup_optimizer(optimizer)
- elif args.qat:
- quantizer = UniformQuantizer(
- model, bits=args.qat, min_size=args.q_min_size)
- return quantizer
-
-
-def load_model(path, strict=False):
- with warnings.catch_warnings():
- warnings.simplefilter("ignore")
- load_from = path
- package = th.load(load_from, 'cpu')
-
- klass = package["klass"]
- args = package["args"]
- kwargs = package["kwargs"]
-
- if strict:
- model = klass(*args, **kwargs)
- else:
- sig = inspect.signature(klass)
- for key in list(kwargs):
- if key not in sig.parameters:
- warnings.warn("Dropping inexistant parameter " + key)
- del kwargs[key]
- model = klass(*args, **kwargs)
-
- state = package["state"]
- training_args = package["training_args"]
- quantizer = get_quantizer(model, training_args)
-
- set_state(model, quantizer, state)
- return model
-
-
-def get_state(model, quantizer):
- if quantizer is None:
- state = {k: p.data.to('cpu') for k, p in model.state_dict().items()}
- else:
- state = quantizer.get_quantized_state()
- buf = io.BytesIO()
- th.save(state, buf)
- state = {'compressed': zlib.compress(buf.getvalue())}
- return state
-
-
-def set_state(model, quantizer, state):
- if quantizer is None:
- model.load_state_dict(state)
- else:
- buf = io.BytesIO(zlib.decompress(state["compressed"]))
- state = th.load(buf, "cpu")
- quantizer.restore_quantized_state(state)
-
- return state
-
-
-def save_state(state, path):
- buf = io.BytesIO()
- th.save(state, buf)
- sig = hashlib.sha256(buf.getvalue()).hexdigest()[:8]
-
- path = path.parent / (path.stem + "-" + sig + path.suffix)
- path.write_bytes(buf.getvalue())
-
-
-def save_model(model, quantizer, training_args, path):
- args, kwargs = model._init_args_kwargs
- klass = model.__class__
-
- state = get_state(model, quantizer)
-
- save_to = path
- package = {
- 'klass': klass,
- 'args': args,
- 'kwargs': kwargs,
- 'state': state,
- 'training_args': training_args,
- }
- th.save(package, save_to)
-
-
-def capture_init(init):
- @functools.wraps(init)
- def __init__(self, *args, **kwargs):
- self._init_args_kwargs = (args, kwargs)
- init(self, *args, **kwargs)
-
- return __init__
diff --git a/demucs/wav.py b/demucs/wav.py
deleted file mode 100644
index a65c3b2ba5aacb1fcab3753f1f85ff7b8db7fc11..0000000000000000000000000000000000000000
--- a/demucs/wav.py
+++ /dev/null
@@ -1,174 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-from collections import OrderedDict
-import hashlib
-import math
-import json
-from pathlib import Path
-
-import julius
-import torch as th
-from torch import distributed
-import torchaudio as ta
-from torch.nn import functional as F
-
-from .audio import convert_audio_channels
-from .compressed import get_musdb_tracks
-
-MIXTURE = "mixture"
-EXT = ".wav"
-
-
-def _track_metadata(track, sources):
- track_length = None
- track_samplerate = None
- for source in sources + [MIXTURE]:
- file = track / f"{source}{EXT}"
- info = ta.info(str(file))
- length = info.num_frames
- if track_length is None:
- track_length = length
- track_samplerate = info.sample_rate
- elif track_length != length:
- raise ValueError(
- f"Invalid length for file {file}: "
- f"expecting {track_length} but got {length}.")
- elif info.sample_rate != track_samplerate:
- raise ValueError(
- f"Invalid sample rate for file {file}: "
- f"expecting {track_samplerate} but got {info.sample_rate}.")
- if source == MIXTURE:
- wav, _ = ta.load(str(file))
- wav = wav.mean(0)
- mean = wav.mean().item()
- std = wav.std().item()
-
- return {"length": length, "mean": mean, "std": std, "samplerate": track_samplerate}
-
-
-def _build_metadata(path, sources):
- meta = {}
- path = Path(path)
- for file in path.iterdir():
- meta[file.name] = _track_metadata(file, sources)
- return meta
-
-
-class Wavset:
- def __init__(
- self,
- root, metadata, sources,
- length=None, stride=None, normalize=True,
- samplerate=44100, channels=2):
- """
- Waveset (or mp3 set for that matter). Can be used to train
- with arbitrary sources. Each track should be one folder inside of `path`.
- The folder should contain files named `{source}.{ext}`.
- Files will be grouped according to `sources` (each source is a list of
- filenames).
-
- Sample rate and channels will be converted on the fly.
-
- `length` is the sample size to extract (in samples, not duration).
- `stride` is how many samples to move by between each example.
- """
- self.root = Path(root)
- self.metadata = OrderedDict(metadata)
- self.length = length
- self.stride = stride or length
- self.normalize = normalize
- self.sources = sources
- self.channels = channels
- self.samplerate = samplerate
- self.num_examples = []
- for name, meta in self.metadata.items():
- track_length = int(self.samplerate * meta['length'] / meta['samplerate'])
- if length is None or track_length < length:
- examples = 1
- else:
- examples = int(math.ceil((track_length - self.length) / self.stride) + 1)
- self.num_examples.append(examples)
-
- def __len__(self):
- return sum(self.num_examples)
-
- def get_file(self, name, source):
- return self.root / name / f"{source}{EXT}"
-
- def __getitem__(self, index):
- for name, examples in zip(self.metadata, self.num_examples):
- if index >= examples:
- index -= examples
- continue
- meta = self.metadata[name]
- num_frames = -1
- offset = 0
- if self.length is not None:
- offset = int(math.ceil(
- meta['samplerate'] * self.stride * index / self.samplerate))
- num_frames = int(math.ceil(
- meta['samplerate'] * self.length / self.samplerate))
- wavs = []
- for source in self.sources:
- file = self.get_file(name, source)
- wav, _ = ta.load(str(file), frame_offset=offset, num_frames=num_frames)
- wav = convert_audio_channels(wav, self.channels)
- wavs.append(wav)
-
- example = th.stack(wavs)
- example = julius.resample_frac(example, meta['samplerate'], self.samplerate)
- if self.normalize:
- example = (example - meta['mean']) / meta['std']
- if self.length:
- example = example[..., :self.length]
- example = F.pad(example, (0, self.length - example.shape[-1]))
- return example
-
-
-def get_wav_datasets(args, samples, sources):
- sig = hashlib.sha1(str(args.wav).encode()).hexdigest()[:8]
- metadata_file = args.metadata / (sig + ".json")
- train_path = args.wav / "train"
- valid_path = args.wav / "valid"
- if not metadata_file.is_file() and args.rank == 0:
- train = _build_metadata(train_path, sources)
- valid = _build_metadata(valid_path, sources)
- json.dump([train, valid], open(metadata_file, "w"))
- if args.world_size > 1:
- distributed.barrier()
- train, valid = json.load(open(metadata_file))
- train_set = Wavset(train_path, train, sources,
- length=samples, stride=args.data_stride,
- samplerate=args.samplerate, channels=args.audio_channels,
- normalize=args.norm_wav)
- valid_set = Wavset(valid_path, valid, [MIXTURE] + sources,
- samplerate=args.samplerate, channels=args.audio_channels,
- normalize=args.norm_wav)
- return train_set, valid_set
-
-
-def get_musdb_wav_datasets(args, samples, sources):
- metadata_file = args.metadata / "musdb_wav.json"
- root = args.musdb / "train"
- if not metadata_file.is_file() and args.rank == 0:
- metadata = _build_metadata(root, sources)
- json.dump(metadata, open(metadata_file, "w"))
- if args.world_size > 1:
- distributed.barrier()
- metadata = json.load(open(metadata_file))
-
- train_tracks = get_musdb_tracks(args.musdb, is_wav=True, subsets=["train"], split="train")
- metadata_train = {name: meta for name, meta in metadata.items() if name in train_tracks}
- metadata_valid = {name: meta for name, meta in metadata.items() if name not in train_tracks}
- train_set = Wavset(root, metadata_train, sources,
- length=samples, stride=args.data_stride,
- samplerate=args.samplerate, channels=args.audio_channels,
- normalize=args.norm_wav)
- valid_set = Wavset(root, metadata_valid, [MIXTURE] + sources,
- samplerate=args.samplerate, channels=args.audio_channels,
- normalize=args.norm_wav)
- return train_set, valid_set
diff --git a/diffq/__init__.py b/diffq/__init__.py
deleted file mode 100644
index 2b997ee4ed99a90cc43db7812383927e6fe1a3e8..0000000000000000000000000000000000000000
--- a/diffq/__init__.py
+++ /dev/null
@@ -1,18 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-# flake8: noqa
-"""
-This package implements different quantization strategies:
-
-- `diffq.uniform.UniformQuantizer`: classic uniform quantization over n bits.
-- `diffq.diffq.DiffQuantizer`: differentiable quantizer based on scaled noise injection.
-
-Also, do check `diffq.base.BaseQuantizer` for the common methods of all Quantizers.
-"""
-
-from .uniform import UniformQuantizer
-from .diffq import DiffQuantizer
diff --git a/diffq/base.py b/diffq/base.py
deleted file mode 100644
index 9bd5276b51fbed3d4b898a45b93479ff19e62a7b..0000000000000000000000000000000000000000
--- a/diffq/base.py
+++ /dev/null
@@ -1,262 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-from dataclasses import dataclass
-from concurrent import futures
-from fnmatch import fnmatch
-from functools import partial
-import io
-import math
-from multiprocessing import cpu_count
-import typing as tp
-import zlib
-
-import torch
-
-
-class BaseQuantizer:
- @dataclass
- class _QuantizedParam:
- name: str
- param: torch.nn.Parameter
- module: torch.nn.Module
- # If a Parameter is used multiple times, `other` can be used
- # to share state between the different Quantizers
- other: tp.Optional[tp.Any]
-
- def __init__(self, model: torch.nn.Module, min_size: float = 0.01, float16: bool = False,
- exclude: tp.Optional[tp.List[str]] = [], detect_bound: bool = True):
- self.model = model
- self.min_size = min_size
- self.float16 = float16
- self.exclude = exclude
- self.detect_bound = detect_bound
- self._quantized = False
- self._pre_handle = self.model.register_forward_pre_hook(self._forward_pre_hook)
- self._post_handle = self.model.register_forward_hook(self._forward_hook)
-
- self._quantized_state = None
- self._qparams = []
- self._float16 = []
- self._others = []
- self._rnns = []
-
- self._saved = []
-
- self._find_params()
-
- def _find_params(self):
- min_params = self.min_size * 2**20 // 4
- previous = {}
- for module_name, module in self.model.named_modules():
- if isinstance(module, torch.nn.RNNBase):
- self._rnns.append(module)
- for name, param in list(module.named_parameters(recurse=False)):
- full_name = f"{module_name}.{name}"
- matched = False
- for pattern in self.exclude:
- if fnmatch(full_name, pattern) or fnmatch(name, pattern):
- matched = True
- break
-
- if param.numel() <= min_params or matched:
- if id(param) in previous:
- continue
- if self.detect_bound:
- previous[id(param)] = None
- if self.float16:
- self._float16.append(param)
- else:
- self._others.append(param)
- else:
- qparam = self._register_param(name, param, module, previous.get(id(param)))
- if self.detect_bound:
- previous[id(param)] = qparam
- self._qparams.append(qparam)
-
- def _register_param(self, name, param, module, other):
- return self.__class__._QuantizedParam(name, param, module, other)
-
- def _forward_pre_hook(self, module, input):
- if self.model.training:
- self._quantized_state = None
- if self._quantized:
- self.unquantize()
- if self._pre_forward_train():
- self._fix_rnns()
- else:
- self.quantize()
-
- def _forward_hook(self, module, input, output):
- if self.model.training:
- if self._post_forward_train():
- self._fix_rnns(flatten=False) # Hacky, next forward will flatten
-
- def quantize(self, save=True):
- """
- Immediately apply quantization to the model parameters.
- If `save` is True, save a copy of the unquantized parameters, that can be
- restored with `unquantize()`.
- """
- if self._quantized:
- return
- if save:
- self._saved = [qp.param.data.to('cpu', copy=True)
- for qp in self._qparams if qp.other is None]
- self.restore_quantized_state(self.get_quantized_state())
- self._quantized = True
- self._fix_rnns()
-
- def unquantize(self):
- """
- Revert a previous call to `quantize()`.
- """
- if not self._quantized:
- raise RuntimeError("Can only be called on a quantized model.")
- if not self._saved:
- raise RuntimeError("Nothing to restore.")
- for qparam in self._qparams:
- if qparam.other is None:
- qparam.param.data[:] = self._saved.pop(0)
- assert len(self._saved) == 0
- self._quantized = False
- self._fix_rnns()
-
- def _pre_forward_train(self) -> bool:
- """
- Called once before each forward for continuous quantization.
- Should return True if parameters were changed.
- """
- return False
-
- def _post_forward_train(self) -> bool:
- """
- Called once after each forward (to restore state for instance).
- Should return True if parameters were changed.
- """
- return False
-
- def _fix_rnns(self, flatten=True):
- """
- To be called after quantization happened to fix RNNs.
- """
- for rnn in self._rnns:
- rnn._flat_weights = [
- (lambda wn: getattr(rnn, wn) if hasattr(rnn, wn) else None)(wn)
- for wn in rnn._flat_weights_names]
- if flatten:
- rnn.flatten_parameters()
-
- def get_quantized_state(self):
- """
- Returns sufficient quantized information to rebuild the model state.
-
- ..Note::
- To achieve maximum compression, you should compress this with
- gzip or other, as quantized weights are not optimally coded!
- """
- if self._quantized_state is None:
- self._quantized_state = self._get_quantized_state()
- return self._quantized_state
-
- def _get_quantized_state(self):
- """
- Actual implementation for `get_quantized_state`.
- """
- float16_params = []
- for p in self._float16:
- q = p.data.half()
- float16_params.append(q)
-
- return {
- "quantized": [self._quantize_param(qparam) for qparam in self._qparams
- if qparam.other is None],
- "float16": float16_params,
- "others": [p.data.clone() for p in self._others],
- }
-
- def _quantize_param(self, qparam: _QuantizedParam) -> tp.Any:
- """
- To be overriden.
- """
- raise NotImplementedError()
-
- def _unquantize_param(self, qparam: _QuantizedParam, quantized: tp.Any) -> torch.Tensor:
- """
- To be overriden.
- """
- raise NotImplementedError()
-
- def restore_quantized_state(self, state) -> None:
- """
- Restore the state of the model from the quantized state.
- """
- for p, q in zip(self._float16, state["float16"]):
- p.data[:] = q.to(p)
-
- for p, q in zip(self._others, state["others"]):
- p.data[:] = q
-
- remaining = list(state["quantized"])
- for qparam in self._qparams:
- if qparam.other is not None:
- # Only unquantize first appearance of nn.Parameter.
- continue
- quantized = remaining.pop(0)
- qparam.param.data[:] = self._unquantize_param(qparam, quantized)
- self._fix_rnns()
-
- def detach(self) -> None:
- """
- Detach from the model, removes hooks and anything else.
- """
- self._pre_handle.remove()
- self._post_handle.remove()
-
- def model_size(self) -> torch.Tensor:
- """
- Returns an estimate of the quantized model size.
- """
- total = torch.tensor(0.)
- for p in self._float16:
- total += 16 * p.numel()
- for p in self._others:
- total += 32 * p.numel()
- return total / 2**20 / 8 # bits to MegaBytes
-
- def true_model_size(self) -> float:
- """
- Return the true quantized model size, in MB, without extra
- compression.
- """
- return self.model_size().item()
-
- def compressed_model_size(self, compress_level=-1, num_workers=8) -> float:
- """
- Return the compressed quantized model size, in MB.
-
- Args:
- compress_level (int): compression level used with zlib,
- see `zlib.compress` for details.
- num_workers (int): will split the final big byte representation in that
- many chunks processed in parallels.
- """
- out = io.BytesIO()
- torch.save(self.get_quantized_state(), out)
- ms = _parallel_compress_len(out.getvalue(), compress_level, num_workers)
- return ms / 2 ** 20
-
-
-def _compress_len(data, compress_level):
- return len(zlib.compress(data, level=compress_level))
-
-
-def _parallel_compress_len(data, compress_level, num_workers):
- num_workers = min(cpu_count(), num_workers)
- chunk_size = int(math.ceil(len(data) / num_workers))
- chunks = [data[offset:offset + chunk_size] for offset in range(0, len(data), chunk_size)]
- with futures.ProcessPoolExecutor(num_workers) as pool:
- return sum(pool.map(partial(_compress_len, compress_level=compress_level), chunks))
diff --git a/diffq/diffq.py b/diffq/diffq.py
deleted file mode 100644
index b475ec7f55227417b014c69b5cf55033182113e1..0000000000000000000000000000000000000000
--- a/diffq/diffq.py
+++ /dev/null
@@ -1,286 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-"""
-Differentiable quantizer based on scaled noise injection.
-"""
-from dataclasses import dataclass
-import math
-import typing as tp
-
-import torch
-
-from .base import BaseQuantizer
-from .uniform import uniform_quantize, uniform_unquantize
-from .utils import simple_repr
-
-
-class DiffQuantizer(BaseQuantizer):
- @dataclass
- class _QuantizedParam(BaseQuantizer._QuantizedParam):
- logit: torch.nn.Parameter
-
- def __init__(self, model: torch.nn.Module, min_size: float = 0.01, float16: bool = False,
- group_size: int = 1, min_bits: float = 2, max_bits: float = 15,
- param="bits", noise="gaussian",
- init_bits: float = 8, extra_bits: float = 0, suffix: str = "_diffq",
- exclude: tp.List[str] = [], detect_bound: bool = True):
- """
- Differentiable quantizer based on scaled noise injection.
- For every parameter `p` in the model, this introduces a number of bits parameter
- `b` with the same dimensions (when group_size = 1).
- Before each forward, `p` is replaced by `p + U`
- with U uniform iid noise with range [-d/2, d/2], with `d` the uniform quantization
- step for `b` bits.
- This noise approximates the quantization noise in a differentiable manner, both
- with respect to the unquantized parameter `p` and the number of bits `b`.
-
- At eveluation (as detected with `model.eval()`), the model is replaced
- by its true quantized version, and restored when going back to training.
-
- When doing actual quantization (for serialization, or evaluation),
- the number of bits is rounded to the nearest integer, and needs to be stored along.
- This will cost a few bits per dimension. To reduce this cost, one can use `group_size`,
- which will use a single noise level for multiple weight entries.
-
- You can use the `DiffQuantizer.model_size` method to get a differentiable estimate of the
- model size in MB. You can then use this estimate as a penalty in your training loss.
-
- Args:
- model (torch.nn.Module): model to quantize
- min_size (float): minimum size in MB of a parameter to be quantized.
- float16 (bool): if a layer is smaller than min_size, should we still do float16?
- group_size (int): weight entries are groupped together to reduce the number
- of noise scales to store. This should divide the size of all parameters
- bigger than min_size.
- min_bits (float): minimal number of bits.
- max_bits (float): maximal number of bits.
- init_bits (float): initial number of bits.
- extra_bits (float): extra bits to add for actual quantization (before roundoff).
- suffix (str): suffix used for the name of the extra noise scale parameters.
- exclude (list[str]): list of patterns used to match parameters to exclude.
- For instance `['bias']` to exclude all bias terms.
- detect_bound (bool): if True, will detect bound parameters and reuse
- the same quantized tensor for both, as well as the same number of bits.
-
- ..Warning::
- You must call `model.training()` and `model.eval()` for `DiffQuantizer` work properly.
-
- """
- self.group_size = group_size
- self.min_bits = min_bits
- self.max_bits = max_bits
- self.init_bits = init_bits
- self.extra_bits = extra_bits
- self.suffix = suffix
- self.param = param
- self.noise = noise
- assert noise in ["gaussian", "uniform"]
- self._optimizer_setup = False
-
- self._min_noise = 1 / (2 ** self.max_bits - 1)
- self._max_noise = 1 / (2 ** self.min_bits - 1)
-
- assert group_size >= 0
- assert min_bits < init_bits < max_bits, \
- "init_bits must be between min_bits and max_bits excluded3"
-
- for name, _ in model.named_parameters():
- if name.endswith(suffix):
- raise RuntimeError("The model already has some noise scales parameters, "
- "maybe you used twice a DiffQuantizer on the same model?.")
-
- super().__init__(model, min_size, float16, exclude, detect_bound)
-
- def _get_bits(self, logit: torch.Tensor):
- if self.param == "noise":
- return torch.log2(1 + 1 / self._get_noise_scale(logit))
- else:
- t = torch.sigmoid(logit)
- return self.max_bits * t + (1 - t) * self.min_bits
-
- def _get_noise_scale(self, logit: torch.Tensor):
- if self.param == "noise":
- t = torch.sigmoid(logit)
- return torch.exp(t * math.log(self._min_noise) + (1 - t) * math.log(self._max_noise))
- else:
- return 1 / (2 ** self._get_bits(logit) - 1)
-
- def _register_param(self, name, param, module, other):
- if other is not None:
- return self.__class__._QuantizedParam(
- name=name, param=param, module=module, logit=other.logit, other=other)
- assert self.group_size == 0 or param.numel() % self.group_size == 0
- # we want the initial number of bits to be init_bits.
- if self.param == "noise":
- noise_scale = 1 / (2 ** self.init_bits - 1)
- t = (math.log(noise_scale) - math.log(self._max_noise)) / (
- math.log(self._min_noise) - math.log(self._max_noise))
- else:
- t = (self.init_bits - self.min_bits) / (self.max_bits - self.min_bits)
- assert 0 < t < 1
- logit = torch.logit(torch.tensor(float(t)))
- assert abs(self._get_bits(logit) - self.init_bits) < 1e-5
- if self.group_size > 0:
- nparam = param.numel() // self.group_size
- else:
- nparam = 1
- logit = torch.nn.Parameter(
- torch.full(
- (nparam,),
- logit,
- device=param.device))
- module.register_parameter(name + self.suffix, logit)
- return self.__class__._QuantizedParam(
- name=name, param=param, module=module, logit=logit, other=None)
-
- def clear_optimizer(self, optimizer: torch.optim.Optimizer):
- params = [qp.logit for qp in self._qparams]
-
- for group in optimizer.param_groups:
- new_params = []
- for q in list(group["params"]):
- matched = False
- for p in params:
- if p is q:
- matched = True
- if not matched:
- new_params.append(q)
- group["params"][:] = new_params
-
- def setup_optimizer(self, optimizer: torch.optim.Optimizer,
- lr: float = 1e-3, **kwargs):
- """
- Setup the optimizer to tune the number of bits. In particular, this will deactivate
- weight decay for the bits parameters.
-
- Args:
- optimizer (torch.Optimizer): optimizer to use.
- lr (float): specific learning rate for the bits parameters. 1e-3
- is perfect for Adam.,w
- kwargs (dict): overrides for other optimization parameters for the bits.
- """
- assert not self._optimizer_setup
- self._optimizer_setup = True
-
- params = [qp.logit for qp in self._qparams]
-
- for group in optimizer.param_groups:
- for q in list(group["params"]):
- for p in params:
- if p is q:
- raise RuntimeError("You should create the optimizer "
- "before the quantizer!")
-
- group = {"params": params, "lr": lr, "weight_decay": 0}
- group.update(kwargs)
- optimizer.add_param_group(group)
-
- def no_optimizer(self):
- """
- Call this if you do not want to use an optimizer.
- """
- self._optimizer_setup = True
-
- def check_unused(self):
- for qparam in self._qparams:
- if qparam.other is not None:
- continue
- grad = qparam.param.grad
- if grad is None or (grad == 0).all():
- if qparam.logit.grad is not None:
- qparam.logit.grad.data.zero_()
-
- def model_size(self, exact=False):
- """
- Differentiable estimate of the model size.
- The size is returned in MB.
-
- If `exact` is True, then the output is no longer differentiable but
- reflect exactly an achievable size, even without compression,
- i.e.same as returned by `naive_model_size()`.
- """
- total = super().model_size()
- subtotal = 0
- for qparam in self._qparams:
- # only count the first appearance of a Parameter
- if qparam.other is not None:
- continue
- bits = self.extra_bits + self._get_bits(qparam.logit)
- if exact:
- bits = bits.round().clamp(1, 15)
- if self.group_size == 0:
- group_size = qparam.param.numel()
- else:
- group_size = self.group_size
- subtotal += group_size * bits.sum()
- subtotal += 2 * 32 # param scale
-
- # Number of bits to represent each number of bits
- bits_bits = math.ceil(math.log2(1 + (bits.max().round().item() - self.min_bits)))
- subtotal += 8 # 8 bits for bits_bits
- subtotal += bits_bits * bits.numel()
-
- subtotal /= 2 ** 20 * 8 # bits -> MegaBytes
- return total + subtotal
-
- def true_model_size(self):
- """
- Naive model size without zlib compression.
- """
- return self.model_size(exact=True).item()
-
- def _pre_forward_train(self):
- if not self._optimizer_setup:
- raise RuntimeError("You must call `setup_optimizer()` on your optimizer "
- "before starting training.")
- for qparam in self._qparams:
- if qparam.other is not None:
- noisy = qparam.other.module._parameters[qparam.other.name]
- else:
- bits = self._get_bits(qparam.logit)[:, None]
- if self.group_size == 0:
- p_flat = qparam.param.view(-1)
- else:
- p_flat = qparam.param.view(-1, self.group_size)
- scale = p_flat.max() - p_flat.min()
- unit = 1 / (2**bits - 1)
- if self.noise == "uniform":
- noise_source = (torch.rand_like(p_flat) - 0.5)
- elif self.noise == "gaussian":
- noise_source = torch.randn_like(p_flat) / 2
- noise = scale * unit * noise_source
- noisy = p_flat + noise
- # We bypass the checks by PyTorch on parameters being leafs
- qparam.module._parameters[qparam.name] = noisy.view_as(qparam.param)
- return True
-
- def _post_forward_train(self):
- for qparam in self._qparams:
- qparam.module._parameters[qparam.name] = qparam.param
- return True
-
- def _quantize_param(self, qparam: _QuantizedParam) -> tp.Any:
- bits = self.extra_bits + self._get_bits(qparam.logit)
- bits = bits.round().clamp(1, 15)[:, None].byte()
- if self.group_size == 0:
- p = qparam.param.data.view(-1)
- else:
- p = qparam.param.data.view(-1, self.group_size)
- levels, scales = uniform_quantize(p, bits)
- return levels, scales, bits
-
- def _unquantize_param(self, qparam: _QuantizedParam, quantized: tp.Any) -> torch.Tensor:
- levels, param_scale, bits = quantized
- return uniform_unquantize(levels, param_scale, bits).view_as(qparam.param.data)
-
- def detach(self):
- super().detach()
- for qparam in self._qparams:
- delattr(qparam.module, qparam.name + self.suffix)
-
- def __repr__(self):
- return simple_repr(self)
diff --git a/diffq/uniform.py b/diffq/uniform.py
deleted file mode 100644
index f61e9129c04caaa33c66f726bf2433d51689cfa5..0000000000000000000000000000000000000000
--- a/diffq/uniform.py
+++ /dev/null
@@ -1,121 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-"""
-Classic uniform quantization over n bits.
-"""
-from typing import Tuple
-import torch
-
-from .base import BaseQuantizer
-from .utils import simple_repr
-
-
-def uniform_quantize(p: torch.Tensor, bits: torch.Tensor = torch.tensor(8.)):
- """
- Quantize the given weights over `bits` bits.
-
- Returns:
- - quantized levels
- - (min, max) range.
-
- """
- assert (bits >= 1).all() and (bits <= 15).all()
- num_levels = (2 ** bits.float()).long()
- mn = p.min().item()
- mx = p.max().item()
- p = (p - mn) / (mx - mn) # put p in [0, 1]
- unit = 1 / (num_levels - 1) # quantization unit
- levels = (p / unit).round()
- if (bits <= 8).all():
- levels = levels.byte()
- else:
- levels = levels.short()
- return levels, (mn, mx)
-
-
-def uniform_unquantize(levels: torch.Tensor, scales: Tuple[float, float],
- bits: torch.Tensor = torch.tensor(8.)):
- """
- Unquantize the weights from the levels and scale. Return a float32 tensor.
- """
- mn, mx = scales
- num_levels = 2 ** bits.float()
- unit = 1 / (num_levels - 1)
- levels = levels.float()
- p = levels * unit # in [0, 1]
- return p * (mx - mn) + mn
-
-
-class UniformQuantizer(BaseQuantizer):
- def __init__(self, model: torch.nn.Module, bits: float = 8., min_size: float = 0.01,
- float16: bool = False, qat: bool = False, exclude=[], detect_bound=True):
- """
- Args:
- model (torch.nn.Module): model to quantize
- bits (float): number of bits to quantize over.
- min_size (float): minimum size in MB of a parameter to be quantized.
- float16 (bool): if a layer is smaller than min_size, should we still do float16?
- qat (bool): perform quantized aware training.
- exclude (list[str]): list of patterns used to match parameters to exclude.
- For instance `['bias']` to exclude all bias terms.
- detect_bound (bool): if True, will detect bound parameters and reuse
- the same quantized tensor for both.
- """
- self.bits = float(bits)
- self.qat = qat
-
- super().__init__(model, min_size, float16, exclude, detect_bound)
-
- def __repr__(self):
- return simple_repr(self, )
-
- def _pre_forward_train(self):
- if self.qat:
- for qparam in self._qparams:
- if qparam.other is not None:
- new_param = qparam.other.module._parameters[qparam.other.name]
- else:
- quantized = self._quantize_param(qparam)
- qvalue = self._unquantize_param(qparam, quantized)
- new_param = qparam.param + (qvalue - qparam.param).detach()
- qparam.module._parameters[qparam.name] = new_param
- return True
- return False
-
- def _post_forward_train(self):
- if self.qat:
- for qparam in self._qparams:
- qparam.module._parameters[qparam.name] = qparam.param
- return True
- return False
-
- def _quantize_param(self, qparam):
- levels, scales = uniform_quantize(qparam.param.data, torch.tensor(self.bits))
- return (levels, scales)
-
- def _unquantize_param(self, qparam, quantized):
- levels, scales = quantized
- return uniform_unquantize(levels, scales, torch.tensor(self.bits))
-
- def model_size(self):
- """
- Non differentiable model size in MB.
- """
- total = super().model_size()
- subtotal = 0
- for qparam in self._qparams:
- if qparam.other is None: # if parameter is bound, count only one copy.
- subtotal += self.bits * qparam.param.numel() + 64 # 2 float for the overall scales
- subtotal /= 2**20 * 8 # bits to MegaBytes
- return total + subtotal
-
- def true_model_size(self):
- """
- Return the true quantized model size, in MB, without extra
- compression.
- """
- return self.model_size().item()
diff --git a/diffq/utils.py b/diffq/utils.py
deleted file mode 100644
index be6ab5253c38564140bc202077292bb99f9f397b..0000000000000000000000000000000000000000
--- a/diffq/utils.py
+++ /dev/null
@@ -1,37 +0,0 @@
-# Copyright (c) Facebook, Inc. and its affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the license found in the
-# LICENSE file in the root directory of this source tree.
-
-import inspect
-from typing import Optional, List
-
-
-def simple_repr(obj, attrs: Optional[List[str]] = None, overrides={}):
- """
- Return a simple representation string for `obj`.
- If `attrs` is not None, it should be a list of attributes to include.
- """
- params = inspect.signature(obj.__class__).parameters
- attrs_repr = []
- if attrs is None:
- attrs = params.keys()
- for attr in attrs:
- display = False
- if attr in overrides:
- value = overrides[attr]
- elif hasattr(obj, attr):
- value = getattr(obj, attr)
- else:
- continue
- if attr in params:
- param = params[attr]
- if param.default is inspect._empty or value != param.default:
- display = True
- else:
- display = True
-
- if display:
- attrs_repr.append(f"{attr}={value}")
- return f"{obj.__class__.__name__}({','.join(attrs_repr)})"
diff --git a/easy_infer.py b/easy_infer.py
deleted file mode 100644
index 5f32bc37e6702cd7288df66059d4a09bc0e3d6aa..0000000000000000000000000000000000000000
--- a/easy_infer.py
+++ /dev/null
@@ -1,1398 +0,0 @@
-import subprocess
-import os
-import sys
-import errno
-import shutil
-import yt_dlp
-from mega import Mega
-import datetime
-import unicodedata
-import torch
-import glob
-import gradio as gr
-import gdown
-import zipfile
-import traceback
-import json
-import mdx
-from mdx_processing_script import get_model_list,id_to_ptm,prepare_mdx,run_mdx
-import requests
-import wget
-import ffmpeg
-import hashlib
-now_dir = os.getcwd()
-sys.path.append(now_dir)
-from unidecode import unidecode
-import re
-import time
-from lib.infer_pack.models_onnx import SynthesizerTrnMsNSFsidM
-from infer.modules.vc.pipeline import Pipeline
-VC = Pipeline
-from lib.infer_pack.models import (
- SynthesizerTrnMs256NSFsid,
- SynthesizerTrnMs256NSFsid_nono,
- SynthesizerTrnMs768NSFsid,
- SynthesizerTrnMs768NSFsid_nono,
-)
-from MDXNet import MDXNetDereverb
-from configs.config import Config
-from infer_uvr5 import _audio_pre_, _audio_pre_new
-from huggingface_hub import HfApi, list_models
-from huggingface_hub import login
-from i18n import I18nAuto
-i18n = I18nAuto()
-from bs4 import BeautifulSoup
-from sklearn.cluster import MiniBatchKMeans
-from dotenv import load_dotenv
-load_dotenv()
-config = Config()
-tmp = os.path.join(now_dir, "TEMP")
-shutil.rmtree(tmp, ignore_errors=True)
-os.environ["TEMP"] = tmp
-weight_root = os.getenv("weight_root")
-weight_uvr5_root = os.getenv("weight_uvr5_root")
-index_root = os.getenv("index_root")
-audio_root = "audios"
-names = []
-for name in os.listdir(weight_root):
- if name.endswith(".pth"):
- names.append(name)
-index_paths = []
-
-global indexes_list
-indexes_list = []
-
-audio_paths = []
-for root, dirs, files in os.walk(index_root, topdown=False):
- for name in files:
- if name.endswith(".index") and "trained" not in name:
- index_paths.append("%s\\%s" % (root, name))
-
-for root, dirs, files in os.walk(audio_root, topdown=False):
- for name in files:
- audio_paths.append("%s/%s" % (root, name))
-
-uvr5_names = []
-for name in os.listdir(weight_uvr5_root):
- if name.endswith(".pth") or "onnx" in name:
- uvr5_names.append(name.replace(".pth", ""))
-
-def calculate_md5(file_path):
- hash_md5 = hashlib.md5()
- with open(file_path, "rb") as f:
- for chunk in iter(lambda: f.read(4096), b""):
- hash_md5.update(chunk)
- return hash_md5.hexdigest()
-
-def format_title(title):
- formatted_title = re.sub(r'[^\w\s-]', '', title)
- formatted_title = formatted_title.replace(" ", "_")
- return formatted_title
-
-def silentremove(filename):
- try:
- os.remove(filename)
- except OSError as e:
- if e.errno != errno.ENOENT:
- raise
-def get_md5(temp_folder):
- for root, subfolders, files in os.walk(temp_folder):
- for file in files:
- if not file.startswith("G_") and not file.startswith("D_") and file.endswith(".pth") and not "_G_" in file and not "_D_" in file:
- md5_hash = calculate_md5(os.path.join(root, file))
- return md5_hash
-
- return None
-
-def find_parent(search_dir, file_name):
- for dirpath, dirnames, filenames in os.walk(search_dir):
- if file_name in filenames:
- return os.path.abspath(dirpath)
- return None
-
-def find_folder_parent(search_dir, folder_name):
- for dirpath, dirnames, filenames in os.walk(search_dir):
- if folder_name in dirnames:
- return os.path.abspath(dirpath)
- return None
-
-
-def delete_large_files(directory_path, max_size_megabytes):
- for filename in os.listdir(directory_path):
- file_path = os.path.join(directory_path, filename)
- if os.path.isfile(file_path):
- size_in_bytes = os.path.getsize(file_path)
- size_in_megabytes = size_in_bytes / (1024 * 1024) # Convert bytes to megabytes
-
- if size_in_megabytes > max_size_megabytes:
- print("###################################")
- print(f"Deleting s*** {filename} (Size: {size_in_megabytes:.2f} MB)")
- os.remove(file_path)
- print("###################################")
-
-def download_from_url(url):
- parent_path = find_folder_parent(".", "pretrained_v2")
- zips_path = os.path.join(parent_path, 'zips')
- print(f"Limit download size in MB {os.getenv('MAX_DOWNLOAD_SIZE')}, duplicate the space for modify the limit")
-
- if url != '':
- print(i18n("Downloading the file: ") + f"{url}")
- if "drive.google.com" in url:
- if "file/d/" in url:
- file_id = url.split("file/d/")[1].split("/")[0]
- elif "id=" in url:
- file_id = url.split("id=")[1].split("&")[0]
- else:
- return None
-
- if file_id:
- os.chdir('./zips')
- result = subprocess.run(["gdown", f"https://drive.google.com/uc?id={file_id}", "--fuzzy"], capture_output=True, text=True, encoding='utf-8')
- if "Too many users have viewed or downloaded this file recently" in str(result.stderr):
- return "too much use"
- if "Cannot retrieve the public link of the file." in str(result.stderr):
- return "private link"
- print(result.stderr)
-
- elif "/blob/" in url:
- os.chdir('./zips')
- url = url.replace("blob", "resolve")
- response = requests.get(url)
- if response.status_code == 200:
- file_name = url.split('/')[-1]
- with open(os.path.join(zips_path, file_name), "wb") as newfile:
- newfile.write(response.content)
- else:
- os.chdir(parent_path)
- elif "mega.nz" in url:
- if "#!" in url:
- file_id = url.split("#!")[1].split("!")[0]
- elif "file/" in url:
- file_id = url.split("file/")[1].split("/")[0]
- else:
- return None
- if file_id:
- m = Mega()
- m.download_url(url, zips_path)
- elif "/tree/main" in url:
- response = requests.get(url)
- soup = BeautifulSoup(response.content, 'html.parser')
- temp_url = ''
- for link in soup.find_all('a', href=True):
- if link['href'].endswith('.zip'):
- temp_url = link['href']
- break
- if temp_url:
- url = temp_url
- url = url.replace("blob", "resolve")
- if "huggingface.co" not in url:
- url = "https://huggingface.co" + url
-
- wget.download(url)
- else:
- print("No .zip file found on the page.")
- elif "cdn.discordapp.com" in url:
- file = requests.get(url)
- if file.status_code == 200:
- name = url.split('/')
- with open(os.path.join(zips_path, name[len(name)-1]), "wb") as newfile:
- newfile.write(file.content)
- else:
- return None
- elif "pixeldrain.com" in url:
- try:
- file_id = url.split("pixeldrain.com/u/")[1]
- os.chdir('./zips')
- print(file_id)
- response = requests.get(f"https://pixeldrain.com/api/file/{file_id}")
- if response.status_code == 200:
- file_name = response.headers.get("Content-Disposition").split('filename=')[-1].strip('";')
- if not os.path.exists(zips_path):
- os.makedirs(zips_path)
- with open(os.path.join(zips_path, file_name), "wb") as newfile:
- newfile.write(response.content)
- os.chdir(parent_path)
- return "downloaded"
- else:
- os.chdir(parent_path)
- return None
- except Exception as e:
- print(e)
- os.chdir(parent_path)
- return None
- else:
- os.chdir('./zips')
- wget.download(url)
-
- #os.chdir('./zips')
- delete_large_files(zips_path, int(os.getenv("MAX_DOWNLOAD_SIZE")))
- os.chdir(parent_path)
- print(i18n("Full download"))
- return "downloaded"
- else:
- return None
-
-class error_message(Exception):
- def __init__(self, mensaje):
- self.mensaje = mensaje
- super().__init__(mensaje)
-
-def get_vc(sid, to_return_protect0, to_return_protect1):
- global n_spk, tgt_sr, net_g, vc, cpt, version
- if sid == "" or sid == []:
- global hubert_model
- if hubert_model is not None:
- print("clean_empty_cache")
- del net_g, n_spk, vc, hubert_model, tgt_sr
- hubert_model = net_g = n_spk = vc = hubert_model = tgt_sr = None
- if torch.cuda.is_available():
- torch.cuda.empty_cache()
- if_f0 = cpt.get("f0", 1)
- version = cpt.get("version", "v1")
- if version == "v1":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs256NSFsid(
- *cpt["config"], is_half=config.is_half
- )
- else:
- net_g = SynthesizerTrnMs256NSFsid_nono(*cpt["config"])
- elif version == "v2":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs768NSFsid(
- *cpt["config"], is_half=config.is_half
- )
- else:
- net_g = SynthesizerTrnMs768NSFsid_nono(*cpt["config"])
- del net_g, cpt
- if torch.cuda.is_available():
- torch.cuda.empty_cache()
- cpt = None
- return (
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- )
- person = "%s/%s" % (weight_root, sid)
- print("loading %s" % person)
- cpt = torch.load(person, map_location="cpu")
- tgt_sr = cpt["config"][-1]
- cpt["config"][-3] = cpt["weight"]["emb_g.weight"].shape[0]
- if_f0 = cpt.get("f0", 1)
- if if_f0 == 0:
- to_return_protect0 = to_return_protect1 = {
- "visible": False,
- "value": 0.5,
- "__type__": "update",
- }
- else:
- to_return_protect0 = {
- "visible": True,
- "value": to_return_protect0,
- "__type__": "update",
- }
- to_return_protect1 = {
- "visible": True,
- "value": to_return_protect1,
- "__type__": "update",
- }
- version = cpt.get("version", "v1")
- if version == "v1":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs256NSFsid(*cpt["config"], is_half=config.is_half)
- else:
- net_g = SynthesizerTrnMs256NSFsid_nono(*cpt["config"])
- elif version == "v2":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs768NSFsid(*cpt["config"], is_half=config.is_half)
- else:
- net_g = SynthesizerTrnMs768NSFsid_nono(*cpt["config"])
- del net_g.enc_q
- print(net_g.load_state_dict(cpt["weight"], strict=False))
- net_g.eval().to(config.device)
- if config.is_half:
- net_g = net_g.half()
- else:
- net_g = net_g.float()
- vc = VC(tgt_sr, config)
- n_spk = cpt["config"][-3]
- return (
- {"visible": True, "maximum": n_spk, "__type__": "update"},
- to_return_protect0,
- to_return_protect1,
- )
-
-def load_downloaded_model(url):
- parent_path = find_folder_parent(".", "pretrained_v2")
- try:
- infos = []
- logs_folders = ['0_gt_wavs','1_16k_wavs','2a_f0','2b-f0nsf','3_feature256','3_feature768']
- zips_path = os.path.join(parent_path, 'zips')
- unzips_path = os.path.join(parent_path, 'unzips')
- weights_path = os.path.join(parent_path, 'weights')
- logs_dir = ""
-
- if os.path.exists(zips_path):
- shutil.rmtree(zips_path)
- if os.path.exists(unzips_path):
- shutil.rmtree(unzips_path)
-
- os.mkdir(zips_path)
- os.mkdir(unzips_path)
-
- download_file = download_from_url(url)
- if not download_file:
- print(i18n("The file could not be downloaded."))
- infos.append(i18n("The file could not be downloaded."))
- yield "\n".join(infos)
- elif download_file == "downloaded":
- print(i18n("It has been downloaded successfully."))
- infos.append(i18n("It has been downloaded successfully."))
- yield "\n".join(infos)
- elif download_file == "too much use":
- raise Exception(i18n("Too many users have recently viewed or downloaded this file"))
- elif download_file == "private link":
- raise Exception(i18n("Cannot get file from this private link"))
-
- for filename in os.listdir(zips_path):
- if filename.endswith(".zip"):
- zipfile_path = os.path.join(zips_path,filename)
- print(i18n("Proceeding with the extraction..."))
- infos.append(i18n("Proceeding with the extraction..."))
- shutil.unpack_archive(zipfile_path, unzips_path, 'zip')
- model_name = os.path.basename(zipfile_path)
- logs_dir = os.path.join(parent_path,'logs', os.path.normpath(str(model_name).replace(".zip","")))
- yield "\n".join(infos)
- else:
- print(i18n("Unzip error."))
- infos.append(i18n("Unzip error."))
- yield "\n".join(infos)
-
- index_file = False
- model_file = False
- D_file = False
- G_file = False
-
- for path, subdirs, files in os.walk(unzips_path):
- for item in files:
- item_path = os.path.join(path, item)
- if not 'G_' in item and not 'D_' in item and item.endswith('.pth'):
- model_file = True
- model_name = item.replace(".pth","")
- logs_dir = os.path.join(parent_path,'logs', model_name)
- if os.path.exists(logs_dir):
- shutil.rmtree(logs_dir)
- os.mkdir(logs_dir)
- if not os.path.exists(weights_path):
- os.mkdir(weights_path)
- if os.path.exists(os.path.join(weights_path, item)):
- os.remove(os.path.join(weights_path, item))
- if os.path.exists(item_path):
- shutil.move(item_path, weights_path)
-
- if not model_file and not os.path.exists(logs_dir):
- os.mkdir(logs_dir)
- for path, subdirs, files in os.walk(unzips_path):
- for item in files:
- item_path = os.path.join(path, item)
- if item.startswith('added_') and item.endswith('.index'):
- index_file = True
- if os.path.exists(item_path):
- if os.path.exists(os.path.join(logs_dir, item)):
- os.remove(os.path.join(logs_dir, item))
- shutil.move(item_path, logs_dir)
- if item.startswith('total_fea.npy') or item.startswith('events.'):
- if os.path.exists(item_path):
- if os.path.exists(os.path.join(logs_dir, item)):
- os.remove(os.path.join(logs_dir, item))
- shutil.move(item_path, logs_dir)
-
-
- result = ""
- if model_file:
- if index_file:
- print(i18n("The model works for inference, and has the .index file."))
- infos.append("\n" + i18n("The model works for inference, and has the .index file."))
- yield "\n".join(infos)
- else:
- print(i18n("The model works for inference, but it doesn't have the .index file."))
- infos.append("\n" + i18n("The model works for inference, but it doesn't have the .index file."))
- yield "\n".join(infos)
-
- if not index_file and not model_file:
- print(i18n("No relevant file was found to upload."))
- infos.append(i18n("No relevant file was found to upload."))
- yield "\n".join(infos)
-
- if os.path.exists(zips_path):
- shutil.rmtree(zips_path)
- if os.path.exists(unzips_path):
- shutil.rmtree(unzips_path)
- os.chdir(parent_path)
- return result
- except Exception as e:
- os.chdir(parent_path)
- if "too much use" in str(e):
- print(i18n("Too many users have recently viewed or downloaded this file"))
- yield i18n("Too many users have recently viewed or downloaded this file")
- elif "private link" in str(e):
- print(i18n("Cannot get file from this private link"))
- yield i18n("Cannot get file from this private link")
- else:
- print(e)
- yield i18n("An error occurred downloading")
- finally:
- os.chdir(parent_path)
-
-def load_dowloaded_dataset(url):
- parent_path = find_folder_parent(".", "pretrained_v2")
- infos = []
- try:
- zips_path = os.path.join(parent_path, 'zips')
- unzips_path = os.path.join(parent_path, 'unzips')
- datasets_path = os.path.join(parent_path, 'datasets')
- audio_extenions =['wav', 'mp3', 'flac', 'ogg', 'opus',
- 'm4a', 'mp4', 'aac', 'alac', 'wma',
- 'aiff', 'webm', 'ac3']
-
- if os.path.exists(zips_path):
- shutil.rmtree(zips_path)
- if os.path.exists(unzips_path):
- shutil.rmtree(unzips_path)
-
- if not os.path.exists(datasets_path):
- os.mkdir(datasets_path)
-
- os.mkdir(zips_path)
- os.mkdir(unzips_path)
-
- download_file = download_from_url(url)
-
- if not download_file:
- print(i18n("An error occurred downloading"))
- infos.append(i18n("An error occurred downloading"))
- yield "\n".join(infos)
- raise Exception(i18n("An error occurred downloading"))
- elif download_file == "downloaded":
- print(i18n("It has been downloaded successfully."))
- infos.append(i18n("It has been downloaded successfully."))
- yield "\n".join(infos)
- elif download_file == "too much use":
- raise Exception(i18n("Too many users have recently viewed or downloaded this file"))
- elif download_file == "private link":
- raise Exception(i18n("Cannot get file from this private link"))
-
- zip_path = os.listdir(zips_path)
- foldername = ""
- for file in zip_path:
- if file.endswith('.zip'):
- file_path = os.path.join(zips_path, file)
- print("....")
- foldername = file.replace(".zip","").replace(" ","").replace("-","_")
- dataset_path = os.path.join(datasets_path, foldername)
- print(i18n("Proceeding with the extraction..."))
- infos.append(i18n("Proceeding with the extraction..."))
- yield "\n".join(infos)
- shutil.unpack_archive(file_path, unzips_path, 'zip')
- if os.path.exists(dataset_path):
- shutil.rmtree(dataset_path)
-
- os.mkdir(dataset_path)
-
- for root, subfolders, songs in os.walk(unzips_path):
- for song in songs:
- song_path = os.path.join(root, song)
- if song.endswith(tuple(audio_extenions)):
- formatted_song_name = format_title(os.path.splitext(song)[0])
- extension = os.path.splitext(song)[1]
- new_song_path = os.path.join(dataset_path, f"{formatted_song_name}{extension}")
- shutil.move(song_path, new_song_path)
- else:
- print(i18n("Unzip error."))
- infos.append(i18n("Unzip error."))
- yield "\n".join(infos)
-
-
-
- if os.path.exists(zips_path):
- shutil.rmtree(zips_path)
- if os.path.exists(unzips_path):
- shutil.rmtree(unzips_path)
-
- print(i18n("The Dataset has been loaded successfully."))
- infos.append(i18n("The Dataset has been loaded successfully."))
- yield "\n".join(infos)
- except Exception as e:
- os.chdir(parent_path)
- if "too much use" in str(e):
- print(i18n("Too many users have recently viewed or downloaded this file"))
- yield i18n("Too many users have recently viewed or downloaded this file")
- elif "private link" in str(e):
- print(i18n("Cannot get file from this private link"))
- yield i18n("Cannot get file from this private link")
- else:
- print(e)
- yield i18n("An error occurred downloading")
- finally:
- os.chdir(parent_path)
-
-def save_model(modelname, save_action):
-
- parent_path = find_folder_parent(".", "pretrained_v2")
- zips_path = os.path.join(parent_path, 'zips')
- dst = os.path.join(zips_path,modelname)
- logs_path = os.path.join(parent_path, 'logs', modelname)
- weights_path = os.path.join(parent_path, 'weights', f"{modelname}.pth")
- save_folder = parent_path
- infos = []
-
- try:
- if not os.path.exists(logs_path):
- raise Exception("No model found.")
-
- if not 'content' in parent_path:
- save_folder = os.path.join(parent_path, 'RVC_Backup')
- else:
- save_folder = '/content/drive/MyDrive/RVC_Backup'
-
- infos.append(i18n("Save model"))
- yield "\n".join(infos)
-
- if not os.path.exists(save_folder):
- os.mkdir(save_folder)
- if not os.path.exists(os.path.join(save_folder, 'ManualTrainingBackup')):
- os.mkdir(os.path.join(save_folder, 'ManualTrainingBackup'))
- if not os.path.exists(os.path.join(save_folder, 'Finished')):
- os.mkdir(os.path.join(save_folder, 'Finished'))
-
- if os.path.exists(zips_path):
- shutil.rmtree(zips_path)
-
- os.mkdir(zips_path)
- added_file = glob.glob(os.path.join(logs_path, "added_*.index"))
- d_file = glob.glob(os.path.join(logs_path, "D_*.pth"))
- g_file = glob.glob(os.path.join(logs_path, "G_*.pth"))
-
- if save_action == i18n("Choose the method"):
- raise Exception("No method choosen.")
-
- if save_action == i18n("Save all"):
- print(i18n("Save all"))
- save_folder = os.path.join(save_folder, 'ManualTrainingBackup')
- shutil.copytree(logs_path, dst)
- else:
- if not os.path.exists(dst):
- os.mkdir(dst)
-
- if save_action == i18n("Save D and G"):
- print(i18n("Save D and G"))
- save_folder = os.path.join(save_folder, 'ManualTrainingBackup')
- if len(d_file) > 0:
- shutil.copy(d_file[0], dst)
- if len(g_file) > 0:
- shutil.copy(g_file[0], dst)
-
- if len(added_file) > 0:
- shutil.copy(added_file[0], dst)
- else:
- infos.append(i18n("Saved without index..."))
-
- if save_action == i18n("Save voice"):
- print(i18n("Save voice"))
- save_folder = os.path.join(save_folder, 'Finished')
- if len(added_file) > 0:
- shutil.copy(added_file[0], dst)
- else:
- infos.append(i18n("Saved without index..."))
-
- yield "\n".join(infos)
- if not os.path.exists(weights_path):
- infos.append(i18n("Saved without inference model..."))
- else:
- shutil.copy(weights_path, dst)
-
- yield "\n".join(infos)
- infos.append("\n" + i18n("This may take a few minutes, please wait..."))
- yield "\n".join(infos)
-
- shutil.make_archive(os.path.join(zips_path,f"{modelname}"), 'zip', zips_path)
- shutil.move(os.path.join(zips_path,f"{modelname}.zip"), os.path.join(save_folder, f'{modelname}.zip'))
-
- shutil.rmtree(zips_path)
- infos.append("\n" + i18n("Model saved successfully"))
- yield "\n".join(infos)
-
- except Exception as e:
- print(e)
- if "No model found." in str(e):
- infos.append(i18n("The model you want to save does not exist, be sure to enter the correct name."))
- else:
- infos.append(i18n("An error occurred saving the model"))
-
- yield "\n".join(infos)
-
-def load_downloaded_backup(url):
- parent_path = find_folder_parent(".", "pretrained_v2")
- try:
- infos = []
- logs_folders = ['0_gt_wavs','1_16k_wavs','2a_f0','2b-f0nsf','3_feature256','3_feature768']
- zips_path = os.path.join(parent_path, 'zips')
- unzips_path = os.path.join(parent_path, 'unzips')
- weights_path = os.path.join(parent_path, 'weights')
- logs_dir = os.path.join(parent_path, 'logs')
-
- if os.path.exists(zips_path):
- shutil.rmtree(zips_path)
- if os.path.exists(unzips_path):
- shutil.rmtree(unzips_path)
-
- os.mkdir(zips_path)
- os.mkdir(unzips_path)
-
- download_file = download_from_url(url)
- if not download_file:
- print(i18n("The file could not be downloaded."))
- infos.append(i18n("The file could not be downloaded."))
- yield "\n".join(infos)
- elif download_file == "downloaded":
- print(i18n("It has been downloaded successfully."))
- infos.append(i18n("It has been downloaded successfully."))
- yield "\n".join(infos)
- elif download_file == "too much use":
- raise Exception(i18n("Too many users have recently viewed or downloaded this file"))
- elif download_file == "private link":
- raise Exception(i18n("Cannot get file from this private link"))
-
- for filename in os.listdir(zips_path):
- if filename.endswith(".zip"):
- zipfile_path = os.path.join(zips_path,filename)
- zip_dir_name = os.path.splitext(filename)[0]
- unzip_dir = unzips_path
- print(i18n("Proceeding with the extraction..."))
- infos.append(i18n("Proceeding with the extraction..."))
- shutil.unpack_archive(zipfile_path, unzip_dir, 'zip')
-
- if os.path.exists(os.path.join(unzip_dir, zip_dir_name)):
- shutil.move(os.path.join(unzip_dir, zip_dir_name), logs_dir)
- else:
- new_folder_path = os.path.join(logs_dir, zip_dir_name)
- os.mkdir(new_folder_path)
- for item_name in os.listdir(unzip_dir):
- item_path = os.path.join(unzip_dir, item_name)
- if os.path.isfile(item_path):
- shutil.move(item_path, new_folder_path)
- elif os.path.isdir(item_path):
- shutil.move(item_path, new_folder_path)
-
- yield "\n".join(infos)
- else:
- print(i18n("Unzip error."))
- infos.append(i18n("Unzip error."))
- yield "\n".join(infos)
-
- result = ""
-
- for filename in os.listdir(unzips_path):
- if filename.endswith(".zip"):
- silentremove(filename)
-
- if os.path.exists(zips_path):
- shutil.rmtree(zips_path)
- if os.path.exists(os.path.join(parent_path, 'unzips')):
- shutil.rmtree(os.path.join(parent_path, 'unzips'))
- print(i18n("The Backup has been uploaded successfully."))
- infos.append("\n" + i18n("The Backup has been uploaded successfully."))
- yield "\n".join(infos)
- os.chdir(parent_path)
- return result
- except Exception as e:
- os.chdir(parent_path)
- if "too much use" in str(e):
- print(i18n("Too many users have recently viewed or downloaded this file"))
- yield i18n("Too many users have recently viewed or downloaded this file")
- elif "private link" in str(e):
- print(i18n("Cannot get file from this private link"))
- yield i18n("Cannot get file from this private link")
- else:
- print(e)
- yield i18n("An error occurred downloading")
- finally:
- os.chdir(parent_path)
-
-def save_to_wav(record_button):
- if record_button is None:
- pass
- else:
- path_to_file=record_button
- new_name = datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S")+'.wav'
- new_path='./audios/'+new_name
- shutil.move(path_to_file,new_path)
- return new_name
-
-
-def change_choices2():
- audio_paths=[]
- for filename in os.listdir("./audios"):
- if filename.endswith(('wav', 'mp3', 'flac', 'ogg', 'opus',
- 'm4a', 'mp4', 'aac', 'alac', 'wma',
- 'aiff', 'webm', 'ac3')):
- audio_paths.append(os.path.join('./audios',filename).replace('\\', '/'))
- return {"choices": sorted(audio_paths), "__type__": "update"}, {"__type__": "update"}
-
-
-
-
-
-def uvr(input_url, output_path, model_name, inp_root, save_root_vocal, paths, save_root_ins, agg, format0, architecture):
- carpeta_a_eliminar = "yt_downloads"
- if os.path.exists(carpeta_a_eliminar) and os.path.isdir(carpeta_a_eliminar):
- for archivo in os.listdir(carpeta_a_eliminar):
- ruta_archivo = os.path.join(carpeta_a_eliminar, archivo)
- if os.path.isfile(ruta_archivo):
- os.remove(ruta_archivo)
- elif os.path.isdir(ruta_archivo):
- shutil.rmtree(ruta_archivo)
-
-
-
- ydl_opts = {
- 'no-windows-filenames': True,
- 'restrict-filenames': True,
- 'extract_audio': True,
- 'format': 'bestaudio',
- 'quiet': True,
- 'no-warnings': True,
- }
-
- try:
- print(i18n("Downloading audio from the video..."))
- with yt_dlp.YoutubeDL(ydl_opts) as ydl:
- info_dict = ydl.extract_info(input_url, download=False)
- formatted_title = format_title(info_dict.get('title', 'default_title'))
- formatted_outtmpl = output_path + '/' + formatted_title + '.wav'
- ydl_opts['outtmpl'] = formatted_outtmpl
- ydl = yt_dlp.YoutubeDL(ydl_opts)
- ydl.download([input_url])
- print(i18n("Audio downloaded!"))
- except Exception as error:
- print(i18n("An error occurred:"), error)
-
- actual_directory = os.path.dirname(__file__)
-
- vocal_directory = os.path.join(actual_directory, save_root_vocal)
- instrumental_directory = os.path.join(actual_directory, save_root_ins)
-
- vocal_formatted = f"vocal_{formatted_title}.wav.reformatted.wav_10.wav"
- instrumental_formatted = f"instrument_{formatted_title}.wav.reformatted.wav_10.wav"
-
- vocal_audio_path = os.path.join(vocal_directory, vocal_formatted)
- instrumental_audio_path = os.path.join(instrumental_directory, instrumental_formatted)
-
- vocal_formatted_mdx = f"{formatted_title}_vocal_.wav"
- instrumental_formatted_mdx = f"{formatted_title}_instrument_.wav"
-
- vocal_audio_path_mdx = os.path.join(vocal_directory, vocal_formatted_mdx)
- instrumental_audio_path_mdx = os.path.join(instrumental_directory, instrumental_formatted_mdx)
-
- if architecture == "VR":
- try:
- print(i18n("Starting audio conversion... (This might take a moment)"))
- inp_root, save_root_vocal, save_root_ins = [x.strip(" ").strip('"').strip("\n").strip('"').strip(" ") for x in [inp_root, save_root_vocal, save_root_ins]]
- usable_files = [os.path.join(inp_root, file)
- for file in os.listdir(inp_root)
- if file.endswith(tuple(sup_audioext))]
-
-
- pre_fun = MDXNetDereverb(15) if model_name == "onnx_dereverb_By_FoxJoy" else (_audio_pre_ if "DeEcho" not in model_name else _audio_pre_new)(
- agg=int(agg),
- model_path=os.path.join(weight_uvr5_root, model_name + ".pth"),
- device=config.device,
- is_half=config.is_half,
- )
-
- try:
- if paths != None:
- paths = [path.name for path in paths]
- else:
- paths = usable_files
-
- except:
- traceback.print_exc()
- paths = usable_files
- print(paths)
- for path in paths:
- inp_path = os.path.join(inp_root, path)
- need_reformat, done = 1, 0
-
- try:
- info = ffmpeg.probe(inp_path, cmd="ffprobe")
- if info["streams"][0]["channels"] == 2 and info["streams"][0]["sample_rate"] == "44100":
- need_reformat = 0
- pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal, format0)
- done = 1
- except:
- traceback.print_exc()
-
- if need_reformat:
- tmp_path = f"{tmp}/{os.path.basename(inp_path)}.reformatted.wav"
- os.system(f"ffmpeg -i {inp_path} -vn -acodec pcm_s16le -ac 2 -ar 44100 {tmp_path} -y")
- inp_path = tmp_path
-
- try:
- if not done:
- pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal, format0)
- print(f"{os.path.basename(inp_path)}->Success")
- except:
- print(f"{os.path.basename(inp_path)}->{traceback.format_exc()}")
- except:
- traceback.print_exc()
- finally:
- try:
- if model_name == "onnx_dereverb_By_FoxJoy":
- del pre_fun.pred.model
- del pre_fun.pred.model_
- else:
- del pre_fun.model
-
- del pre_fun
- return i18n("Finished"), vocal_audio_path, instrumental_audio_path
- except: traceback.print_exc()
-
- if torch.cuda.is_available(): torch.cuda.empty_cache()
-
- elif architecture == "MDX":
- try:
- print(i18n("Starting audio conversion... (This might take a moment)"))
- inp_root, save_root_vocal, save_root_ins = [x.strip(" ").strip('"').strip("\n").strip('"').strip(" ") for x in [inp_root, save_root_vocal, save_root_ins]]
-
- usable_files = [os.path.join(inp_root, file)
- for file in os.listdir(inp_root)
- if file.endswith(tuple(sup_audioext))]
- try:
- if paths != None:
- paths = [path.name for path in paths]
- else:
- paths = usable_files
-
- except:
- traceback.print_exc()
- paths = usable_files
- print(paths)
- invert=True
- denoise=True
- use_custom_parameter=True
- dim_f=2048
- dim_t=256
- n_fft=7680
- use_custom_compensation=True
- compensation=1.025
- suffix = "vocal_" #@param ["Vocals", "Drums", "Bass", "Other"]{allow-input: true}
- suffix_invert = "instrument_" #@param ["Instrumental", "Drumless", "Bassless", "Instruments"]{allow-input: true}
- print_settings = True # @param{type:"boolean"}
- onnx = id_to_ptm(model_name)
- compensation = compensation if use_custom_compensation or use_custom_parameter else None
- mdx_model = prepare_mdx(onnx,use_custom_parameter, dim_f, dim_t, n_fft, compensation=compensation)
-
-
- for path in paths:
- #inp_path = os.path.join(inp_root, path)
- suffix_naming = suffix if use_custom_parameter else None
- diff_suffix_naming = suffix_invert if use_custom_parameter else None
- run_mdx(onnx, mdx_model, path, format0, diff=invert,suffix=suffix_naming,diff_suffix=diff_suffix_naming,denoise=denoise)
-
- if print_settings:
- print()
- print('[MDX-Net_Colab settings used]')
- print(f'Model used: {onnx}')
- print(f'Model MD5: {mdx.MDX.get_hash(onnx)}')
- print(f'Model parameters:')
- print(f' -dim_f: {mdx_model.dim_f}')
- print(f' -dim_t: {mdx_model.dim_t}')
- print(f' -n_fft: {mdx_model.n_fft}')
- print(f' -compensation: {mdx_model.compensation}')
- print()
- print('[Input file]')
- print('filename(s): ')
- for filename in paths:
- print(f' -{filename}')
- print(f"{os.path.basename(filename)}->Success")
- except:
- traceback.print_exc()
- finally:
- try:
- del mdx_model
- return i18n("Finished"), vocal_audio_path_mdx, instrumental_audio_path_mdx
- except: traceback.print_exc()
-
- print("clean_empty_cache")
-
- if torch.cuda.is_available(): torch.cuda.empty_cache()
-sup_audioext = {'wav', 'mp3', 'flac', 'ogg', 'opus',
- 'm4a', 'mp4', 'aac', 'alac', 'wma',
- 'aiff', 'webm', 'ac3'}
-
-def load_downloaded_audio(url):
- parent_path = find_folder_parent(".", "pretrained_v2")
- try:
- infos = []
- audios_path = os.path.join(parent_path, 'audios')
- zips_path = os.path.join(parent_path, 'zips')
-
- if not os.path.exists(audios_path):
- os.mkdir(audios_path)
-
- download_file = download_from_url(url)
- if not download_file:
- print(i18n("The file could not be downloaded."))
- infos.append(i18n("The file could not be downloaded."))
- yield "\n".join(infos)
- elif download_file == "downloaded":
- print(i18n("It has been downloaded successfully."))
- infos.append(i18n("It has been downloaded successfully."))
- yield "\n".join(infos)
- elif download_file == "too much use":
- raise Exception(i18n("Too many users have recently viewed or downloaded this file"))
- elif download_file == "private link":
- raise Exception(i18n("Cannot get file from this private link"))
-
- for filename in os.listdir(zips_path):
- item_path = os.path.join(zips_path, filename)
- if item_path.split('.')[-1] in sup_audioext:
- if os.path.exists(item_path):
- shutil.move(item_path, audios_path)
-
- result = ""
- print(i18n("Audio files have been moved to the 'audios' folder."))
- infos.append(i18n("Audio files have been moved to the 'audios' folder."))
- yield "\n".join(infos)
-
- os.chdir(parent_path)
- return result
- except Exception as e:
- os.chdir(parent_path)
- if "too much use" in str(e):
- print(i18n("Too many users have recently viewed or downloaded this file"))
- yield i18n("Too many users have recently viewed or downloaded this file")
- elif "private link" in str(e):
- print(i18n("Cannot get file from this private link"))
- yield i18n("Cannot get file from this private link")
- else:
- print(e)
- yield i18n("An error occurred downloading")
- finally:
- os.chdir(parent_path)
-
-
-class error_message(Exception):
- def __init__(self, mensaje):
- self.mensaje = mensaje
- super().__init__(mensaje)
-
-def get_vc(sid, to_return_protect0, to_return_protect1):
- global n_spk, tgt_sr, net_g, vc, cpt, version
- if sid == "" or sid == []:
- global hubert_model
- if hubert_model is not None:
- print("clean_empty_cache")
- del net_g, n_spk, vc, hubert_model, tgt_sr
- hubert_model = net_g = n_spk = vc = hubert_model = tgt_sr = None
- if torch.cuda.is_available():
- torch.cuda.empty_cache()
- if_f0 = cpt.get("f0", 1)
- version = cpt.get("version", "v1")
- if version == "v1":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs256NSFsid(
- *cpt["config"], is_half=config.is_half
- )
- else:
- net_g = SynthesizerTrnMs256NSFsid_nono(*cpt["config"])
- elif version == "v2":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs768NSFsid(
- *cpt["config"], is_half=config.is_half
- )
- else:
- net_g = SynthesizerTrnMs768NSFsid_nono(*cpt["config"])
- del net_g, cpt
- if torch.cuda.is_available():
- torch.cuda.empty_cache()
- cpt = None
- return (
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- {"visible": False, "__type__": "update"},
- )
- person = "%s/%s" % (weight_root, sid)
- print("loading %s" % person)
- cpt = torch.load(person, map_location="cpu")
- tgt_sr = cpt["config"][-1]
- cpt["config"][-3] = cpt["weight"]["emb_g.weight"].shape[0]
- if_f0 = cpt.get("f0", 1)
- if if_f0 == 0:
- to_return_protect0 = to_return_protect1 = {
- "visible": False,
- "value": 0.5,
- "__type__": "update",
- }
- else:
- to_return_protect0 = {
- "visible": True,
- "value": to_return_protect0,
- "__type__": "update",
- }
- to_return_protect1 = {
- "visible": True,
- "value": to_return_protect1,
- "__type__": "update",
- }
- version = cpt.get("version", "v1")
- if version == "v1":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs256NSFsid(*cpt["config"], is_half=config.is_half)
- else:
- net_g = SynthesizerTrnMs256NSFsid_nono(*cpt["config"])
- elif version == "v2":
- if if_f0 == 1:
- net_g = SynthesizerTrnMs768NSFsid(*cpt["config"], is_half=config.is_half)
- else:
- net_g = SynthesizerTrnMs768NSFsid_nono(*cpt["config"])
- del net_g.enc_q
- print(net_g.load_state_dict(cpt["weight"], strict=False))
- net_g.eval().to(config.device)
- if config.is_half:
- net_g = net_g.half()
- else:
- net_g = net_g.float()
- vc = VC(tgt_sr, config)
- n_spk = cpt["config"][-3]
- return (
- {"visible": True, "maximum": n_spk, "__type__": "update"},
- to_return_protect0,
- to_return_protect1,
- )
-
-def update_model_choices(select_value):
- model_ids = get_model_list()
- model_ids_list = list(model_ids)
- if select_value == "VR":
- return {"choices": uvr5_names, "__type__": "update"}
- elif select_value == "MDX":
- return {"choices": model_ids_list, "__type__": "update"}
-
-def download_model():
- gr.Markdown(value="# " + i18n("Download Model"))
- gr.Markdown(value=i18n("It is used to download your inference models."))
- with gr.Row():
- model_url=gr.Textbox(label=i18n("Url:"))
- with gr.Row():
- download_model_status_bar=gr.Textbox(label=i18n("Status:"))
- with gr.Row():
- download_button=gr.Button(i18n("Download"))
- download_button.click(fn=load_downloaded_model, inputs=[model_url], outputs=[download_model_status_bar])
-
-def download_backup():
- gr.Markdown(value="# " + i18n("Download Backup"))
- gr.Markdown(value=i18n("It is used to download your training backups."))
- with gr.Row():
- model_url=gr.Textbox(label=i18n("Url:"))
- with gr.Row():
- download_model_status_bar=gr.Textbox(label=i18n("Status:"))
- with gr.Row():
- download_button=gr.Button(i18n("Download"))
- download_button.click(fn=load_downloaded_backup, inputs=[model_url], outputs=[download_model_status_bar])
-
-def update_dataset_list(name):
- new_datasets = []
- for foldername in os.listdir("./datasets"):
- if "." not in foldername:
- new_datasets.append(os.path.join(find_folder_parent(".","pretrained"),"datasets",foldername))
- return gr.Dropdown.update(choices=new_datasets)
-
-def download_dataset(trainset_dir4):
- gr.Markdown(value="# " + i18n("Download Dataset"))
- gr.Markdown(value=i18n("Download the dataset with the audios in a compatible format (.wav/.flac) to train your model."))
- with gr.Row():
- dataset_url=gr.Textbox(label=i18n("Url:"))
- with gr.Row():
- load_dataset_status_bar=gr.Textbox(label=i18n("Status:"))
- with gr.Row():
- load_dataset_button=gr.Button(i18n("Download"))
- load_dataset_button.click(fn=load_dowloaded_dataset, inputs=[dataset_url], outputs=[load_dataset_status_bar])
- load_dataset_status_bar.change(update_dataset_list, dataset_url, trainset_dir4)
-
-def download_audio():
- gr.Markdown(value="# " + i18n("Download Audio"))
- gr.Markdown(value=i18n("Download audios of any format for use in inference (recommended for mobile users)."))
- with gr.Row():
- audio_url=gr.Textbox(label=i18n("Url:"))
- with gr.Row():
- download_audio_status_bar=gr.Textbox(label=i18n("Status:"))
- with gr.Row():
- download_button2=gr.Button(i18n("Download"))
- download_button2.click(fn=load_downloaded_audio, inputs=[audio_url], outputs=[download_audio_status_bar])
-
-def youtube_separator():
- gr.Markdown(value="# " + i18n("Separate YouTube tracks"))
- gr.Markdown(value=i18n("Download audio from a YouTube video and automatically separate the vocal and instrumental tracks"))
- with gr.Row():
- input_url = gr.inputs.Textbox(label=i18n("Enter the YouTube link:"))
- output_path = gr.Textbox(
- label=i18n("Enter the path of the audio folder to be processed (copy it from the address bar of the file manager):"),
- value=os.path.abspath(os.getcwd()).replace('\\', '/') + "/yt_downloads",
- visible=False,
- )
- advanced_settings_checkbox = gr.Checkbox(
- value=False,
- label=i18n("Advanced Settings"),
- interactive=True,
- )
- with gr.Row(label = i18n("Advanced Settings"), visible=False, variant='compact') as advanced_settings:
- with gr.Column():
- model_select = gr.Radio(
- label=i18n("Model Architecture:"),
- choices=["VR", "MDX"],
- value="VR",
- interactive=True,
- )
- model_choose = gr.Dropdown(label=i18n("Model: (Be aware that in some models the named vocal will be the instrumental)"),
- choices=uvr5_names,
- value="HP5_only_main_vocal"
- )
- with gr.Row():
- agg = gr.Slider(
- minimum=0,
- maximum=20,
- step=1,
- label=i18n("Vocal Extraction Aggressive"),
- value=10,
- interactive=True,
- )
- with gr.Row():
- opt_vocal_root = gr.Textbox(
- label=i18n("Specify the output folder for vocals:"), value="audios",
- )
- opt_ins_root = gr.Textbox(
- label=i18n("Specify the output folder for accompaniment:"), value="audio-others",
- )
- dir_wav_input = gr.Textbox(
- label=i18n("Enter the path of the audio folder to be processed:"),
- value=((os.getcwd()).replace('\\', '/') + "/yt_downloads"),
- visible=False,
- )
- format0 = gr.Radio(
- label=i18n("Export file format"),
- choices=["wav", "flac", "mp3", "m4a"],
- value="wav",
- visible=False,
- interactive=True,
- )
- wav_inputs = gr.File(
- file_count="multiple", label=i18n("You can also input audio files in batches. Choose one of the two options. Priority is given to reading from the folder."),
- visible=False,
- )
- model_select.change(
- fn=update_model_choices,
- inputs=model_select,
- outputs=model_choose,
- )
- with gr.Row():
- vc_output4 = gr.Textbox(label=i18n("Status:"))
- vc_output5 = gr.Audio(label=i18n("Vocal"), type='filepath')
- vc_output6 = gr.Audio(label=i18n("Instrumental"), type='filepath')
- with gr.Row():
- but2 = gr.Button(i18n("Download and Separate"))
- but2.click(
- uvr,
- [
- input_url,
- output_path,
- model_choose,
- dir_wav_input,
- opt_vocal_root,
- wav_inputs,
- opt_ins_root,
- agg,
- format0,
- model_select
- ],
- [vc_output4, vc_output5, vc_output6],
- )
- def toggle_advanced_settings(checkbox):
- return {"visible": checkbox, "__type__": "update"}
-
- advanced_settings_checkbox.change(
- fn=toggle_advanced_settings,
- inputs=[advanced_settings_checkbox],
- outputs=[advanced_settings]
- )
-
-
-def get_bark_voice():
- mensaje = """
-v2/en_speaker_0 English Male
-v2/en_speaker_1 English Male
-v2/en_speaker_2 English Male
-v2/en_speaker_3 English Male
-v2/en_speaker_4 English Male
-v2/en_speaker_5 English Male
-v2/en_speaker_6 English Male
-v2/en_speaker_7 English Male
-v2/en_speaker_8 English Male
-v2/en_speaker_9 English Female
-v2/zh_speaker_0 Chinese (Simplified) Male
-v2/zh_speaker_1 Chinese (Simplified) Male
-v2/zh_speaker_2 Chinese (Simplified) Male
-v2/zh_speaker_3 Chinese (Simplified) Male
-v2/zh_speaker_4 Chinese (Simplified) Female
-v2/zh_speaker_5 Chinese (Simplified) Male
-v2/zh_speaker_6 Chinese (Simplified) Female
-v2/zh_speaker_7 Chinese (Simplified) Female
-v2/zh_speaker_8 Chinese (Simplified) Male
-v2/zh_speaker_9 Chinese (Simplified) Female
-v2/fr_speaker_0 French Male
-v2/fr_speaker_1 French Female
-v2/fr_speaker_2 French Female
-v2/fr_speaker_3 French Male
-v2/fr_speaker_4 French Male
-v2/fr_speaker_5 French Female
-v2/fr_speaker_6 French Male
-v2/fr_speaker_7 French Male
-v2/fr_speaker_8 French Male
-v2/fr_speaker_9 French Male
-v2/de_speaker_0 German Male
-v2/de_speaker_1 German Male
-v2/de_speaker_2 German Male
-v2/de_speaker_3 German Female
-v2/de_speaker_4 German Male
-v2/de_speaker_5 German Male
-v2/de_speaker_6 German Male
-v2/de_speaker_7 German Male
-v2/de_speaker_8 German Female
-v2/de_speaker_9 German Male
-v2/hi_speaker_0 Hindi Female
-v2/hi_speaker_1 Hindi Female
-v2/hi_speaker_2 Hindi Male
-v2/hi_speaker_3 Hindi Female
-v2/hi_speaker_4 Hindi Female
-v2/hi_speaker_5 Hindi Male
-v2/hi_speaker_6 Hindi Male
-v2/hi_speaker_7 Hindi Male
-v2/hi_speaker_8 Hindi Male
-v2/hi_speaker_9 Hindi Female
-v2/it_speaker_0 Italian Male
-v2/it_speaker_1 Italian Male
-v2/it_speaker_2 Italian Female
-v2/it_speaker_3 Italian Male
-v2/it_speaker_4 Italian Male
-v2/it_speaker_5 Italian Male
-v2/it_speaker_6 Italian Male
-v2/it_speaker_7 Italian Female
-v2/it_speaker_8 Italian Male
-v2/it_speaker_9 Italian Female
-v2/ja_speaker_0 Japanese Female
-v2/ja_speaker_1 Japanese Female
-v2/ja_speaker_2 Japanese Male
-v2/ja_speaker_3 Japanese Female
-v2/ja_speaker_4 Japanese Female
-v2/ja_speaker_5 Japanese Female
-v2/ja_speaker_6 Japanese Male
-v2/ja_speaker_7 Japanese Female
-v2/ja_speaker_8 Japanese Female
-v2/ja_speaker_9 Japanese Female
-v2/ko_speaker_0 Korean Female
-v2/ko_speaker_1 Korean Male
-v2/ko_speaker_2 Korean Male
-v2/ko_speaker_3 Korean Male
-v2/ko_speaker_4 Korean Male
-v2/ko_speaker_5 Korean Male
-v2/ko_speaker_6 Korean Male
-v2/ko_speaker_7 Korean Male
-v2/ko_speaker_8 Korean Male
-v2/ko_speaker_9 Korean Male
-v2/pl_speaker_0 Polish Male
-v2/pl_speaker_1 Polish Male
-v2/pl_speaker_2 Polish Male
-v2/pl_speaker_3 Polish Male
-v2/pl_speaker_4 Polish Female
-v2/pl_speaker_5 Polish Male
-v2/pl_speaker_6 Polish Female
-v2/pl_speaker_7 Polish Male
-v2/pl_speaker_8 Polish Male
-v2/pl_speaker_9 Polish Female
-v2/pt_speaker_0 Portuguese Male
-v2/pt_speaker_1 Portuguese Male
-v2/pt_speaker_2 Portuguese Male
-v2/pt_speaker_3 Portuguese Male
-v2/pt_speaker_4 Portuguese Male
-v2/pt_speaker_5 Portuguese Male
-v2/pt_speaker_6 Portuguese Male
-v2/pt_speaker_7 Portuguese Male
-v2/pt_speaker_8 Portuguese Male
-v2/pt_speaker_9 Portuguese Male
-v2/ru_speaker_0 Russian Male
-v2/ru_speaker_1 Russian Male
-v2/ru_speaker_2 Russian Male
-v2/ru_speaker_3 Russian Male
-v2/ru_speaker_4 Russian Male
-v2/ru_speaker_5 Russian Female
-v2/ru_speaker_6 Russian Female
-v2/ru_speaker_7 Russian Male
-v2/ru_speaker_8 Russian Male
-v2/ru_speaker_9 Russian Female
-v2/es_speaker_0 Spanish Male
-v2/es_speaker_1 Spanish Male
-v2/es_speaker_2 Spanish Male
-v2/es_speaker_3 Spanish Male
-v2/es_speaker_4 Spanish Male
-v2/es_speaker_5 Spanish Male
-v2/es_speaker_6 Spanish Male
-v2/es_speaker_7 Spanish Male
-v2/es_speaker_8 Spanish Female
-v2/es_speaker_9 Spanish Female
-v2/tr_speaker_0 Turkish Male
-v2/tr_speaker_1 Turkish Male
-v2/tr_speaker_2 Turkish Male
-v2/tr_speaker_3 Turkish Male
-v2/tr_speaker_4 Turkish Female
-v2/tr_speaker_5 Turkish Female
-v2/tr_speaker_6 Turkish Male
-v2/tr_speaker_7 Turkish Male
-v2/tr_speaker_8 Turkish Male
-v2/tr_speaker_9 Turkish Male
- """
-# Dividir el mensaje en líneas
- lineas = mensaje.split("\n")
- datos_deseados = []
- for linea in lineas:
- partes = linea.split("\t")
- if len(partes) == 3:
- clave, _, genero = partes
- datos_deseados.append(f"{clave}-{genero}")
-
- return datos_deseados
-
-
-def get_edge_voice():
- completed_process = subprocess.run(['edge-tts',"-l"], capture_output=True, text=True)
- lines = completed_process.stdout.strip().split("\n")
- data = []
- current_entry = {}
- for line in lines:
- if line.startswith("Name: "):
- if current_entry:
- data.append(current_entry)
- current_entry = {"Name": line.split(": ")[1]}
- elif line.startswith("Gender: "):
- current_entry["Gender"] = line.split(": ")[1]
- if current_entry:
- data.append(current_entry)
- tts_voice = []
- for entry in data:
- name = entry["Name"]
- gender = entry["Gender"]
- formatted_entry = f'{name}-{gender}'
- tts_voice.append(formatted_entry)
- return tts_voice
-
-
-#print(set_tts_voice)
diff --git a/environment_dml.yaml b/environment_dml.yaml
deleted file mode 100644
index 0fb3f222554eb01acce5313bf81cee4179edf0af..0000000000000000000000000000000000000000
--- a/environment_dml.yaml
+++ /dev/null
@@ -1,186 +0,0 @@
-name: pydml
-channels:
- - pytorch
- - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main
- - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/
- - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
- - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge/
- - defaults
- - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/fastai/
- - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/
- - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/bioconda/
-dependencies:
- - abseil-cpp=20211102.0=hd77b12b_0
- - absl-py=1.3.0=py310haa95532_0
- - aiohttp=3.8.3=py310h2bbff1b_0
- - aiosignal=1.2.0=pyhd3eb1b0_0
- - async-timeout=4.0.2=py310haa95532_0
- - attrs=22.1.0=py310haa95532_0
- - blas=1.0=mkl
- - blinker=1.4=py310haa95532_0
- - bottleneck=1.3.5=py310h9128911_0
- - brotli=1.0.9=h2bbff1b_7
- - brotli-bin=1.0.9=h2bbff1b_7
- - brotlipy=0.7.0=py310h2bbff1b_1002
- - bzip2=1.0.8=he774522_0
- - c-ares=1.19.0=h2bbff1b_0
- - ca-certificates=2023.05.30=haa95532_0
- - cachetools=4.2.2=pyhd3eb1b0_0
- - certifi=2023.5.7=py310haa95532_0
- - cffi=1.15.1=py310h2bbff1b_3
- - charset-normalizer=2.0.4=pyhd3eb1b0_0
- - click=8.0.4=py310haa95532_0
- - colorama=0.4.6=py310haa95532_0
- - contourpy=1.0.5=py310h59b6b97_0
- - cryptography=39.0.1=py310h21b164f_0
- - cycler=0.11.0=pyhd3eb1b0_0
- - fonttools=4.25.0=pyhd3eb1b0_0
- - freetype=2.12.1=ha860e81_0
- - frozenlist=1.3.3=py310h2bbff1b_0
- - giflib=5.2.1=h8cc25b3_3
- - glib=2.69.1=h5dc1a3c_2
- - google-auth=2.6.0=pyhd3eb1b0_0
- - google-auth-oauthlib=0.4.4=pyhd3eb1b0_0
- - grpc-cpp=1.48.2=hf108199_0
- - grpcio=1.48.2=py310hf108199_0
- - gst-plugins-base=1.18.5=h9e645db_0
- - gstreamer=1.18.5=hd78058f_0
- - icu=58.2=ha925a31_3
- - idna=3.4=py310haa95532_0
- - intel-openmp=2023.1.0=h59b6b97_46319
- - jpeg=9e=h2bbff1b_1
- - kiwisolver=1.4.4=py310hd77b12b_0
- - krb5=1.19.4=h5b6d351_0
- - lerc=3.0=hd77b12b_0
- - libbrotlicommon=1.0.9=h2bbff1b_7
- - libbrotlidec=1.0.9=h2bbff1b_7
- - libbrotlienc=1.0.9=h2bbff1b_7
- - libclang=14.0.6=default_hb5a9fac_1
- - libclang13=14.0.6=default_h8e68704_1
- - libdeflate=1.17=h2bbff1b_0
- - libffi=3.4.4=hd77b12b_0
- - libiconv=1.16=h2bbff1b_2
- - libogg=1.3.5=h2bbff1b_1
- - libpng=1.6.39=h8cc25b3_0
- - libprotobuf=3.20.3=h23ce68f_0
- - libtiff=4.5.0=h6c2663c_2
- - libuv=1.44.2=h2bbff1b_0
- - libvorbis=1.3.7=he774522_0
- - libwebp=1.2.4=hbc33d0d_1
- - libwebp-base=1.2.4=h2bbff1b_1
- - libxml2=2.10.3=h0ad7f3c_0
- - libxslt=1.1.37=h2bbff1b_0
- - lz4-c=1.9.4=h2bbff1b_0
- - markdown=3.4.1=py310haa95532_0
- - markupsafe=2.1.1=py310h2bbff1b_0
- - matplotlib=3.7.1=py310haa95532_1
- - matplotlib-base=3.7.1=py310h4ed8f06_1
- - mkl=2023.1.0=h8bd8f75_46356
- - mkl-service=2.4.0=py310h2bbff1b_1
- - mkl_fft=1.3.6=py310h4ed8f06_1
- - mkl_random=1.2.2=py310h4ed8f06_1
- - multidict=6.0.2=py310h2bbff1b_0
- - munkres=1.1.4=py_0
- - numexpr=2.8.4=py310h2cd9be0_1
- - numpy=1.24.3=py310h055cbcc_1
- - numpy-base=1.24.3=py310h65a83cf_1
- - oauthlib=3.2.2=py310haa95532_0
- - openssl=1.1.1t=h2bbff1b_0
- - packaging=23.0=py310haa95532_0
- - pandas=1.5.3=py310h4ed8f06_0
- - pcre=8.45=hd77b12b_0
- - pillow=9.4.0=py310hd77b12b_0
- - pip=23.0.1=py310haa95532_0
- - ply=3.11=py310haa95532_0
- - protobuf=3.20.3=py310hd77b12b_0
- - pyasn1=0.4.8=pyhd3eb1b0_0
- - pyasn1-modules=0.2.8=py_0
- - pycparser=2.21=pyhd3eb1b0_0
- - pyjwt=2.4.0=py310haa95532_0
- - pyopenssl=23.0.0=py310haa95532_0
- - pyparsing=3.0.9=py310haa95532_0
- - pyqt=5.15.7=py310hd77b12b_0
- - pyqt5-sip=12.11.0=py310hd77b12b_0
- - pysocks=1.7.1=py310haa95532_0
- - python=3.10.11=h966fe2a_2
- - python-dateutil=2.8.2=pyhd3eb1b0_0
- - pytorch-mutex=1.0=cpu
- - pytz=2022.7=py310haa95532_0
- - pyyaml=6.0=py310h2bbff1b_1
- - qt-main=5.15.2=he8e5bd7_8
- - qt-webengine=5.15.9=hb9a9bb5_5
- - qtwebkit=5.212=h2bbfb41_5
- - re2=2022.04.01=hd77b12b_0
- - requests=2.29.0=py310haa95532_0
- - requests-oauthlib=1.3.0=py_0
- - rsa=4.7.2=pyhd3eb1b0_1
- - setuptools=67.8.0=py310haa95532_0
- - sip=6.6.2=py310hd77b12b_0
- - six=1.16.0=pyhd3eb1b0_1
- - sqlite=3.41.2=h2bbff1b_0
- - tbb=2021.8.0=h59b6b97_0
- - tensorboard=2.10.0=py310haa95532_0
- - tensorboard-data-server=0.6.1=py310haa95532_0
- - tensorboard-plugin-wit=1.8.1=py310haa95532_0
- - tk=8.6.12=h2bbff1b_0
- - toml=0.10.2=pyhd3eb1b0_0
- - tornado=6.2=py310h2bbff1b_0
- - tqdm=4.65.0=py310h9909e9c_0
- - typing_extensions=4.5.0=py310haa95532_0
- - tzdata=2023c=h04d1e81_0
- - urllib3=1.26.16=py310haa95532_0
- - vc=14.2=h21ff451_1
- - vs2015_runtime=14.27.29016=h5e58377_2
- - werkzeug=2.2.3=py310haa95532_0
- - wheel=0.38.4=py310haa95532_0
- - win_inet_pton=1.1.0=py310haa95532_0
- - xz=5.4.2=h8cc25b3_0
- - yaml=0.2.5=he774522_0
- - yarl=1.8.1=py310h2bbff1b_0
- - zlib=1.2.13=h8cc25b3_0
- - zstd=1.5.5=hd43e919_0
- - pip:
- - antlr4-python3-runtime==4.8
- - appdirs==1.4.4
- - audioread==3.0.0
- - bitarray==2.7.4
- - cython==0.29.35
- - decorator==5.1.1
- - fairseq==0.12.2
- - faiss-cpu==1.7.4
- - filelock==3.12.0
- - hydra-core==1.0.7
- - jinja2==3.1.2
- - joblib==1.2.0
- - lazy-loader==0.2
- - librosa==0.10.0.post2
- - llvmlite==0.40.0
- - lxml==4.9.2
- - mpmath==1.3.0
- - msgpack==1.0.5
- - networkx==3.1
- - noisereduce==2.0.1
- - numba==0.57.0
- - omegaconf==2.0.6
- - opencv-python==4.7.0.72
- - pooch==1.6.0
- - portalocker==2.7.0
- - pysimplegui==4.60.5
- - pywin32==306
- - pyworld==0.3.3
- - regex==2023.5.5
- - sacrebleu==2.3.1
- - scikit-learn==1.2.2
- - scipy==1.10.1
- - sounddevice==0.4.6
- - soundfile==0.12.1
- - soxr==0.3.5
- - sympy==1.12
- - tabulate==0.9.0
- - threadpoolctl==3.1.0
- - torch==2.0.0
- - torch-directml==0.2.0.dev230426
- - torchaudio==2.0.1
- - torchvision==0.15.1
- - wget==3.2
-prefix: D:\ProgramData\anaconda3_\envs\pydml
diff --git a/extract_locale.py b/extract_locale.py
deleted file mode 100644
index a4ff5ea3ddd7c612c640544099ab98a861b8fe35..0000000000000000000000000000000000000000
--- a/extract_locale.py
+++ /dev/null
@@ -1,34 +0,0 @@
-import json
-import re
-
-# Define regular expression patterns
-pattern = r"""i18n\([\s\n\t]*(["'][^"']+["'])[\s\n\t]*\)"""
-
-# Initialize the dictionary to store key-value pairs
-data = {}
-
-
-def process(fn: str):
- global data
- with open(fn, "r", encoding="utf-8") as f:
- contents = f.read()
- matches = re.findall(pattern, contents)
- for key in matches:
- key = eval(key)
- print("extract:", key)
- data[key] = key
-
-
-print("processing infer-web.py")
-process("infer-web.py")
-
-print("processing gui_v0.py")
-process("gui_v0.py")
-
-print("processing gui_v1.py")
-process("gui_v1.py")
-
-# Save as a JSON file
-with open("./i18n/en_US.json", "w", encoding="utf-8") as f:
- json.dump(data, f, ensure_ascii=False, indent=4)
- f.write("\n")
diff --git a/formantshiftcfg/f2m.txt b/formantshiftcfg/f2m.txt
deleted file mode 100644
index 40356a80ce7dd7a893bca233a41306525193f2f0..0000000000000000000000000000000000000000
--- a/formantshiftcfg/f2m.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-1.0
-0.8
\ No newline at end of file
diff --git a/formantshiftcfg/m2f.txt b/formantshiftcfg/m2f.txt
deleted file mode 100644
index fa69b52dc8e29db4697401ef2766dd0b6c6f4b47..0000000000000000000000000000000000000000
--- a/formantshiftcfg/m2f.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-1.0
-1.2
\ No newline at end of file
diff --git a/formantshiftcfg/random.txt b/formantshiftcfg/random.txt
deleted file mode 100644
index 427be5c80412098ecec082b3a06e867ddc9a7ba2..0000000000000000000000000000000000000000
--- a/formantshiftcfg/random.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-32.0
-9.8
\ No newline at end of file
diff --git a/get-pip.py b/get-pip.py
deleted file mode 100644
index cf68a2b3426decf01bde021a50c47a2115af303a..0000000000000000000000000000000000000000
--- a/get-pip.py
+++ /dev/null
@@ -1,32657 +0,0 @@
-#!/usr/bin/env python
-#
-# Hi There!
-#
-# You may be wondering what this giant blob of binary data here is, you might
-# even be worried that we're up to something nefarious (good for you for being
-# paranoid!). This is a base85 encoding of a zip file, this zip file contains
-# an entire copy of pip (version 23.2.1).
-#
-# Pip is a thing that installs packages, pip itself is a package that someone
-# might want to install, especially if they're looking to run this get-pip.py
-# script. Pip has a lot of code to deal with the security of installing
-# packages, various edge cases on various platforms, and other such sort of
-# "tribal knowledge" that has been encoded in its code base. Because of this
-# we basically include an entire copy of pip inside this blob. We do this
-# because the alternatives are attempt to implement a "minipip" that probably
-# doesn't do things correctly and has weird edge cases, or compress pip itself
-# down into a single file.
-#
-# If you're wondering how this is created, it is generated using
-# `scripts/generate.py` in https://github.com/pypa/get-pip.
-
-import sys
-
-this_python = sys.version_info[:2]
-min_version = (3, 7)
-if this_python < min_version:
- message_parts = [
- "This script does not work on Python {}.{}".format(*this_python),
- "The minimum supported Python version is {}.{}.".format(*min_version),
- "Please use https://bootstrap.pypa.io/pip/{}.{}/get-pip.py instead.".format(*this_python),
- ]
- print("ERROR: " + " ".join(message_parts))
- sys.exit(1)
-
-
-import os.path
-import pkgutil
-import shutil
-import tempfile
-import argparse
-import importlib
-from base64 import b85decode
-
-
-def include_setuptools(args):
- """
- Install setuptools only if absent and not excluded.
- """
- cli = not args.no_setuptools
- env = not os.environ.get("PIP_NO_SETUPTOOLS")
- absent = not importlib.util.find_spec("setuptools")
- return cli and env and absent
-
-
-def include_wheel(args):
- """
- Install wheel only if absent and not excluded.
- """
- cli = not args.no_wheel
- env = not os.environ.get("PIP_NO_WHEEL")
- absent = not importlib.util.find_spec("wheel")
- return cli and env and absent
-
-
-def determine_pip_install_arguments():
- pre_parser = argparse.ArgumentParser()
- pre_parser.add_argument("--no-setuptools", action="store_true")
- pre_parser.add_argument("--no-wheel", action="store_true")
- pre, args = pre_parser.parse_known_args()
-
- args.append("pip")
-
- if include_setuptools(pre):
- args.append("setuptools")
-
- if include_wheel(pre):
- args.append("wheel")
-
- return ["install", "--upgrade", "--force-reinstall"] + args
-
-
-def monkeypatch_for_cert(tmpdir):
- """Patches `pip install` to provide default certificate with the lowest priority.
-
- This ensures that the bundled certificates are used unless the user specifies a
- custom cert via any of pip's option passing mechanisms (config, env-var, CLI).
-
- A monkeypatch is the easiest way to achieve this, without messing too much with
- the rest of pip's internals.
- """
- from pip._internal.commands.install import InstallCommand
-
- # We want to be using the internal certificates.
- cert_path = os.path.join(tmpdir, "cacert.pem")
- with open(cert_path, "wb") as cert:
- cert.write(pkgutil.get_data("pip._vendor.certifi", "cacert.pem"))
-
- install_parse_args = InstallCommand.parse_args
-
- def cert_parse_args(self, args):
- if not self.parser.get_default_values().cert:
- # There are no user provided cert -- force use of bundled cert
- self.parser.defaults["cert"] = cert_path # calculated above
- return install_parse_args(self, args)
-
- InstallCommand.parse_args = cert_parse_args
-
-
-def bootstrap(tmpdir):
- monkeypatch_for_cert(tmpdir)
-
- # Execute the included pip and use it to install the latest pip and
- # setuptools from PyPI
- from pip._internal.cli.main import main as pip_entry_point
- args = determine_pip_install_arguments()
- sys.exit(pip_entry_point(args))
-
-
-def main():
- tmpdir = None
- try:
- # Create a temporary working directory
- tmpdir = tempfile.mkdtemp()
-
- # Unpack the zipfile into the temporary directory
- pip_zip = os.path.join(tmpdir, "pip.zip")
- with open(pip_zip, "wb") as fp:
- fp.write(b85decode(DATA.replace(b"\n", b"")))
-
- # Add the zipfile to sys.path so that we can import it
- sys.path.insert(0, pip_zip)
-
- # Run the bootstrap
- bootstrap(tmpdir=tmpdir)
- finally:
- # Clean up our temporary working directory
- if tmpdir:
- shutil.rmtree(tmpdir, ignore_errors=True)
-
-
-DATA = b"""
-P)h>@6aWAK2ml*O_EvJ33*7hs003nH000jF003}la4%n9X>MtBUtcb8c|B0UO2j}6z0X&KUUXrd;wrc
-n6ubz6s0VM$QfAw<4YV^ulDhQoop$MlK*;0ehK(>
-3ZJA0oQV`^+*aO7_tw^Cd$4zs{Pl#j>6{|X*AaQ6!2wJ?w>%d+2&1X4Rc!^r6h-hMtH_d5{IF3D`nKTt~p1QY-O00;mZO7>Q7_pjHy0RRA2
-0{{RI0001RX>c!JUu|J&ZeL$6aCu!)OK;mS48HqU5b43r;JP^vOMxACEp{6QLy+m1h%E`C9MAjpBNe-
-8r;{H19{ebpf{zJ27j)n8%0=-6Z#elILRo@w9oRWWbO{z8ujDS!QAC@3T%nJCf;1rX6ghzu#Z}R@K&*?Hgj1WFD91+adaM4G`4Xs@*hA^t@nbDYdL)-aOjsW~3}QVVby(8=@7U$
-Fzj5Y{w!2hUUH`?e9j7WDA;>-1aos>7j{2$~BfyL8p@__Y98dsP#Bs7^lWF
-=e_gr;(4^?am?Cp93+7b-!?~nb}-$cPSR1zckA*zNp!)$;YjlZrfn&RWNM}=QA7*cb8A{(9@{5!vBfq
-rEMoeu5FvJZngI@N#4#(2v$WnMGCAVD?b9t8W^qDfcFBe5ZZF%dPAPaq#>ikclG~yPvCg`JUGb_W2#PdCXxx}7!|T*xc9qdnTILbO-nAJaF2
-~0snMFDU<%E01X4*yW9@|}F2;vY~;0|XQR000O88%p+8eg`F$&;kGeqy+!~6#xJLaA|NaUte%(a4
-m9mZf<3AUtcb8d3{vDZrd;nz56RT_b?l9y`mj3AXtV0MT+&(WJ!7$x|Xq_^eh*q`
-qYNbl$TgcX!{RW4b=Vw*pI`moV*K|DJ2bY*KQViviHGglIK{X_)>pN=IEr427|<0g`vfCSX-CrF6hnx-
-fU6^LzLVM{GttvQ!RX(K-@qvQ<9nZh3{TwCd*xxj~wep|+d4YrpRGd3uJ(;$x#MJ^wO(dX9-I(W~SOL
-|!j@ev4#PBd+t2O-3Y4TDlA%@&y9h}l?d7(gvc*a&O+atWdOv5|
-XtFg8N1I1Eg2~6T^Prn{|GZSIw2~Ql9c?>!a3=lwO6eT!TZzV{RAoH`=gPAEk0OKF^-L_LxAV)%Ld>V
-rC7Ea!84dqJ@cSb~%=6Dm=^V^deci#%k)qhs`k`mikNs;GRv|TRB1+w&XWHK8?pSmvO+Mn5HP0Rg&
-0e2!{O&s!2A%Oz`W5|6)QOoeMptG0vVbf-p%MA<(l*rGUrRG$G|nf0000U0RR9D
-0001RX>c!ac`kH$aAjmAj&U%1)EvFVxRjSzi=C>J@cM
-k87yJyz4~-Qcqlg}hXv}1CF`fEox?~SG{pae%Dy$pBG>tnWs3{>FohpTZSG@fe-hAmws@4PFv7Mv`H@
-JnAXTbgKqwrl)IWaYE>+%OsO9KQH0000802@m7R+hDJp-}+<06hW#02u%P0B~t=FJEbHbY*gGVQep7U
-ukY>bYEXCaCvo+F;B!W42Adn3hP*|5~K?fa1xA6Cs^1JI)&D4jnX98E~x(=w{RdN$P(+xdH(X;aUMbE
-La7HDOJ;>ViJroJQOYSq=f31Z#UCgsvZ;PjisC7~V50}YW@1zhN!C_4fs|i^>lX7r-W?|$V(y(g0ZOD
-x-5bTWf^iasXM`rih^?Sk#%z{jZl{Ri-7?Gn9_p
-NH(fR_VZQx#ZustU5xCHVj%1=)fT*F;XSi#wiQR~iuoy}(RFp&L9pfC#Zn^7Axz>2yIKB7|@~y3-1&J5eC&FySna4hw0fjd92G^LTzc+Br>7Y7w1=({
-s_3<|LwzLQl3jT^=CKlyadUgD7M{+)3>-rRJjjOO9KQH0000802@m7Rtz49V_OUW00Srh02%-Q0B~t=
-FJEbHbY*gGVQepAb!lv5UuAA~E^v9(T1#`|xDmeVS714ZB@>eSIHgphB=gYhxH9p$W;~m0sZ?Bwghq@
-hk_(WwwJ!hnbT%XT~X$2UELOWbx^D5}
-p)=7nt84rjpQ!t=bvqBu6SXjxf*{)}V#v6kjnleUMl*qKLJw7ma)>Zw|O-`#U0v?-c6x#d+}i#X$=E%t?3;qCzPO{p3XDn-l0g8$MLf}@FhxjzhJPffk$LZTb=
-tRL0mAd`8KB>SS|Ny1Wz!%10ZP4l!(Z9X~Qr(M}5e1M{2V-3vl>e`}|vFvt@s535m*|M}OlVSM$)RrHcBrimd6?lFPUdh
-^8oI-}L;caqLRJjDa?_Dr07Ysf#%z>QWYbSDW3_SKrT&dAFG`Lt`@W9KJiJ}-Jm
-Eim0UQMILLA#<&5?}IiA5v%!>tEItSETqsiWmt%EBta_NRXj{H*Zo{ba+L0f#Cr>zURR@B*qHa1TLRl
-QIY3XdTuN;Q8cY|sQ{2jC4o$vPgD13HO~sl#?~l?=&A}cMFP(CNl(yMsR`-t2i}7DFz8rYgPveC_)gi
-?sXaiv@_U|jtx7a74!l@<;4JHe05F%Q2)SdHLgXxn>Gh!i1WT2K^_-Mqe1LMOvU4R{fH+QfQ%eQYa2d
-+e#MFwQ*oaQwvhGC2wTnRW_zJ##J9Pw*x1bE%az6lfqS#7Kz)e-Rnn7GhG_W5G{(q)4xvM*rJ>eb1rMlGrLDy?OC^}{4osLlt4f7K8F}Z|`B#E1o*9RQ|@+2V@Bv`<7P)h{}C>a!R4k{Eil{;q0l?#-&mQ~4}M0|c2#OI;L{3Tudz_N!_rY+hTGzghD(#5kNVHprZaZYt##W$uR8%mb^&)N6ivKk8Fogh
-BMr8%*?0wS)XN@6p#nApa&Y-w9Ew#Zu@h&NSzS79U`3ykgq8X+X_$OD4A`np9UQ(OZ&?G>pd7)u100h
-6&Ehk$^P1yyq9_uwBi4bItZ;{YLK4idID%pU;f7}zm-6NU3Bg;MsQ)C_Xl%pd#APfelK6ZX)4MevarN3gu`&(XOy?+-ilBrvl6uD3dNNZ)`pUd=i?WZkc;yu4_~oaIcdwK6<&R*Ivfg2cB}&44
-buBuR0s5klsH#FHwVF%6r=l3b;v1Sm=o@?fr!Fer2uDL9L&_isoatz297jX@o{A}`XCC6WOfkP0%87K
-kvdJZNsFYvO_uCPfDQ#!T$k!x23L!YQl05fIp!Qum#J6eLAwB~uW~Tzhl*@BpO*0iZn3-W@i=nr-W|(
-11w{K!f#O~7gF*~{#IxcX?lmI`bj@X}m2N^O|Q*fI&p?b#N0ES?Pkc+y}Zj6*0Av01gLhWTd
-nOTbhdhE1BKPS3Fg`Z@s&2l@TrvgBPRJC~P2NN1QqdwS}`bs=5XEmp~mF78qqjMEpthH9w@9Bbi4O^<
-&W#%iuEa|8!D0@I9@+SrhW~?;jIcMk1?8@}gUVbulb{fRenF5C)Hq^fLi3aX)oB9={Aw5B1Eya}ud)zI$Kgq)lD
-w*#4Ftggw^aT0%+Vu2)HvSC$KTnIVg4EflgOXhv&oh3ZSz1L>6#^bnq(m1-OmeK*vFp=M92_79T`!l}
-{9`kKpLiSkSXPgGNTXzCM!m$>YmKH1A`kiHiQ*43y-)7dhX|M$wzbh0!*8wWuO
-$+?!aMXV))oSMbZk=4=^~Bzkn$82y9L)OTJZ?WBoqw*oo)B@x+ciJET=1kF<^8cqPG
-P!?R*vWNM|ELa#g+A5(FI=e?5HVwlmM86Sq%vDSn84<7Nu4Cy@v^93Go0~&XH@{(?4R;W-+{wnaSX4h
-dBLaWBKHJuX_r9tZX^)!C5M>y}CCk2Bg3Q180j_`3MbCnUAON=wR_Mw>=B(2!qdobEOu2v5=yT@shGM
-~r3jQ4Lc*S5nc8W7-2G(!r^M~cFZ6m}#W&xX`N#98l}tUwm`Ct`*ss)D%~d2{jaf3BD8RZT}pLU*I=(
-}#C|8y|~WONGYELk8FDK9$3V(nS{-09xll!z%G^#&nYYK%@=^l2j(@kWt-j^soOk{KTUk>+MW2)n^^6
-(IL-fyvERX*?qG*p`hD|5y$?@0$n)X&O2H;^6Ex)SV+1jP-txm+iOw9lzzJAF$`cMda)C%T
-GVJkVYGtLqIROwK@kZ{Ay>IU@jDOX%0SdYm|x;oqbm2$vloyp_(hz1t7I43OljOG#o7wOvTf?rAeARa
-|#tj9{cl%YbU1ah!CbL`!H33}dzr
-htU}qX&Y?3r~m~9(#^Nq?X+Q|_9G!Gbecu}-EupvSfdppnjX=t2xj3q;=s^aZd#RHI3bE@&IndzQQe?
-i1`zO-;MmiOM@SbDofi_1tz~EAd#Gh=@ohyX!HEeD{|0MK8X+k#$FHr^$7_}l_w`+ZnbT?no-_f_d2^
-gF__@%r^IIH%GBQ!NI72pmqqVd1vE@1Pr*4|@_{B@EF0PV~*Do$#zj*ila-FfeFvbZm^^FGfkZ#J7flbgK~uVgNF>Y#FaF`LaUF7%-+Dl7KV>@&S?9zU2OZ+>URZm
-08I^H`XRZB-mZDJ|^~e)wBFx(RzKvAh|7nx7WpE4{G`@lqRnzbUOQa+zItGP;bDTa~9p6_;}JQPNqll
-{?c=cqexYp>wOMvQqd?a(Phwky}+65WS0HZFSa?+{nDh^+sm;N5$kqW|%M-jMb-&VrJWYFY;ULN#F04
-%D&c_;;kb)4@Ign6Q{aT8=KTs))4rLN4~GJJ9cF{|Jba5iQjiDJrX0$TIOnOF^e8sbtn^X)T$NFj-8@
-{iD(+L$w!^1W||6QX|+Kfkl2FcySN}PQI%LV?h@~meaT}{!YWRZ`NhSX?_PZK;&t-(w{Ko2ub;lU!un
-ZJX>5qe<=~GOsoIK!+!4%fY?Ln9d#;VG76M;4bMf#m^kaD;@PQA1r)*v2LSj&^GbPMkK6><66k7}t3G
-%k;6qC2p4udo4tT?R?rHN8dg)qrSbuz1WRSnNFs+5(4TFfe%EoKWbTh8VSp>k7KDv@TRHLsjAy~-W$1
-1NTW?#JpWLXRdK6RW#F?EzNxpE#>lp)
-L@KQmY%OvdbHSvR#Q(wVAd@e}J8Z3r!je`je)Ck^oa=V6;$d%XlO!@K+b%*1V2d^Xy2w4ltjxNEf#-3
-%Z{ALUeFZ1U3)_(q;J7d`IZmt%WR2RXZX+EXcUxBd?R0*?FT5;q^X!cf+#1h3DP+kJ#Eet+Auqb=xQF
-Q9DDq=$BF)ebs7G3HsErkC)iV2`(78&*QQLjTPOCZkd?wy2ag;d-6k?}x1rJg}=7ORhL$$#ZPN^$zNj
-Tg>9AVKS|J*h^19BgTg-Si7jbyU#zk3OeHjX#w6z)PBwmsR4&u*u~na&oyv#6o7)N(dOcowdnwS-4$gvNc5Z?Za7Vbu|?4jtqNxFtz=&^dnjT11v;4IL1ID{P8VIaktI_HeD
-ph^a6s9Mm}XTh}^EIel%nssby5GkriNRV6AM)!D*Xygb1)d3!pB5HAL~sf^1LseG+ybyersu{iS!=M*
-kuF}3F4jbb^91EN5$b*Ak}P;HI_0()yqv%I|AL85vcWASBqD&-~0$E7x=R_5}LkN*5*=wa8h^Qz7UIU
-fvi%EVSL^kBCir+nM7d*!60Km<7sPqu|i+!T_ZXPIzO$7*Xs4&En>KIlwV0X?HObwzqXrbaTfl$l{}e1zjN>Bi_ss8)0_r(d{QhFWq$@a3>XM?j}$soOk6IBI
-&u#!le>AE~2k_cxYGN!FNPI=l!F}8{4xeyA%X@jh$HDXDyZ`<-I6shZfA>e12}YZck^uA^W39p#_)p2
-?1tXA~?YD!vm?bE50NhSCj^BGD}h;>RuQ2#i7i&_fqLqRTWi}nLKk*4+C{cI`FT+ETMNbO%(&2ZV})a
-A$64|l(d)5`Or`KCEg)Hd?>D=q(YqtM3teKWMm{m^@+;W!hw$?$yfP(7znro3iP!k!L^00BCW;vx+q
-NX+TzEJ+U|VJoPE$|UH24jzygsZkM7^8isYA9!ZY7WyC6|sYS73jlUte`zO3Aa%^$)d*#Z|nEMSSVQD
-}>o;@grJ5^0LOK=tx&yb%(S`XAOR9&=~f75q}ZKF-<~5a7#-pFZMQ;&TtGsIz)hE2gRavlZHtiLkZvY
-{ZwSV!k~w`La8T+vQ`*YDPlDZwC{UD=Zn-1a&Y^0ko|)NIqvml?~u{qy;lD}oo_g+kf^F0msufq*K--n%A)TfOG-cds<|N9Wt{C3jrdGQS^DtTw`*s)69PYV+`VKuz?ioy
-k*jhrh$Xa2Iy{pQ=Fyl3ea31xSnpRs%F+~5Fr=M|@FUGJj7EPZA#cIXKycEe0A9=P<3{}8VzrnB+QX`x)Je2|u-Qb7SFaxVx`#8w*J1
-?3mE?gP}g#VDb`lB4rk&&9(}DF1N9{N5@jA)J!i*ogX3G6MTP6My3*$(1bVg6Qn
-!M<@5_Si=GPt~X7Kslz*{=cXBHm-h3x${9v=mfe)LP!=9;k=?-iw;ej0gk3T6$LlFQ7_SU{0H>s^A5X
-RmB11KbUVj#A&NVnbCr5iDb9=9evaaM%=xCes3hP=|4Q^Ujqt`09VfB&Qhe3F5AVAI-HESy7v%BPx6i
-L$CGVeKUA;`6J$?54s&BwuR;+;#PRW5^x1{c7xm^-ibAT-SjIplI?u)PTHiPyg{W)cffQhkmQ6|=OrXGd}tA1a}6Ww-9cMttV?p{AAlI_b6x~JiL+-AL7nso0=65n;wTcxPwCEg;}^VA%A;aqSj&R;3?n
-$c0ZvFlx;`;?_X60p-s5Yv@=La!0Vlg7lsaAl+m^~pwj*M_>Vn$>pu@K3X0Djh9+SIsLaDjd@l=_sol
-#DA(LX;(&l94T@u&$%E?2W2XHVCECWWn9$hLc>O!&WNLBrg8u29`Jo7%R#nbcs~q;;hx;bqPP3bll?F3jeP|ga5#Al)jQ3x_KVI8q`TLdzllbIEagHuxU>s~JyP!>yR
-(CO7kVK@`0wzoO$a#7#)7?Z2S^$Vg-rb~G}mtJs_M#JrkmB3!{49w2^?@W>4O@Xw-K*(-ps(;%|1m^9fopm0d{)eQjvRPV
-r~Jt6SOhBZZ{nvUJ`3Uh3yHs$nK=`gHg#{t|7t9X)aIbjC#7x3p*|N?nxw$cSpgj43U}16A`8*Z>wH@
-*JIFIqUwr8gv07It?g!0&E*`rRiiI@qfLgyEvU*n3S<0rR{Oc{i2pONRc`?tYpf4{_vWiNOMZXYsa^e0Yj~!AIh~K}y6eSo2|#uEhcCW-@|c1Fd2C0wmF
-F~CKcT-qpqrVVbJF0DTl(C`SoE2_aTcP#tr!~U0bTI+ZeW_@dBeEnnCt@+J2Z)Znf|DN9VPual~~t1!
-M7Ri90X)lJFnw6b_0|XQR000O88%p+8N-^Lqfe-)y4>kY*9smFUaA|NaUukZ1WpZv|Y%gPPZf0p`b#h^JX>V>
-WaCyxe{cqbg_ILjkoQk3{=BVp#INS{Q?y|-01xu44>Bk1c&=iY~d66Z7lVj+@>r)(#x6-zXbBX-C4u;?6q0EG38$n6SIy;6Y0g76B>mk4(a3
-Az+XULh7tiTpO>Y*)yXrCcqf05G>~x8f2|UvYz)r4dd%BIH<^2+07sE1_*#v`w|Z}kB{^Hh@FT3LE8#LtQ(<>_cJ>^o;uiL5>%Da%wyb#Pq-!YY
-%>F8_Rbesb~o`tWj4om+=Dx4b%oC&1f-JJv!i>~fx~jpQ+4G=lG&^@=O1BqBEPBo?(_vlr}o1&
-~%ro(_Hyc?uhh5W)a|2P38`IUEdrzBqq-`Y!(I_n<_B4A=^31vJ}T))9{gTeItQ;h4c<
-I{KN7gy60+_>dZfeZk4u;N(>+Vz5c0DZiJ0~ITlzG5oWRnXW(@@Sx!Oo&=7?vK~gt4Xi{Y5*S4^AYK~
-F8M+%#e!D6JG=Pl_-qo~X2ngC=~dTzRq-|ZEK*Kuu1`NqCxH?b*Y9Vagse76HfPg(D`b(A?R#K>v`N7
-8t=>TLx;(v%4Wr(ko=xqt_|x$fEd~3M&T<#@Cp26z1qDiY@o9Q>b$T+5FRo6eS3oUM9cem7<`>d!za#
-ecJDlfy#iIwGj?Yd{;0`oJ
-v79`GqA_1rBFZ!PA1}gjxKvt$7(AEQnl1w&d1YE7$DmB>n=^9_R|c&Tw}
-!J2(Po}*xkJlnHU@+B}XE5NdWABr|e2plrk{(YdSPlbX2z}F!67#PzcAARBse$2-fogfN;l@*2+T3RE
-*(apucRs~@SFbeB8#J*qno}~p>v>CWpB>*8UFqna3py*>G3OE9kQN#it#3h%jq*QEQY}gJW3~T|pqR?
-MyyNd1~UIAhtmL&a0vw0XTQN&fKBb0qC69HSht~&H68MYZ0sWKB)2z(f^H$%fl(9YQN7%>IgkeG;pW`
->?@)bP_VRO4;7>OH`^S&d_%B5>ua=--9NL;N;kEiX7^KpewYC=wGJBJ?5_Dn1C&9~zyS59l9v2_6jRd
-Z$6?j8KWhm+qMaAOpQ$>>mTsA%lM@LAdBB!{EQca8xfK^tw(w!pF_378?46MkeRG$1t&c!J{3%7`7ZB
-&Unz|dO=v%<>;tQnA!ROdbh~HaDDx)UDdyU&8SOG(%6n^kye9CS!^K
-!DX=<5VSWi-2<<9aEWiRVF+h7HK={K?*i1@EkUp%0Vj3w}1Oc{E;Ds9K$Cszzogp!)z>g{x*t`*wJ~X
-TVyv{z;u@IgqI#*SN)E>F11%XGcnE~3vNrGo!voxs^O+NrbWEBK4v49N32w77?UK}zsQ
-N=Oj;@NTptBVXdGzP>ANMJL_En|!d<2tJ)e>BHbtH?QdDg4rSbG0ccEY*;QgZdqqX$=uELywNT3G?QS
-i4v{IKlXh3K_Bd`B6{BuI8XwS8dX5KHOCz>wZd$gK_Y<2fWF^91lIL;;1NuSwAw$clQM(|3^{BI-Qlp
-a(|^+ZLf%J~^}t#C)nCvcJZX?`c>99=#1{$1v>hfYykwP37I#REE}G!+EpVb%5Gs$n6Jnict8pjrP$7
-fiXT(}r_NxOAb7HmmjWYGK!+O_43lXj19v<|SFn}8Du|w}4zVgs@kwSjV(}oRC(vBf_-d
-Gcgg)FJZF2L-tQd4c~#az0_pvfehRM2LC4Z5TQZVU;BuA$|WAvucW+m8doIVA?JDQmGdJwf1cRm41n?
-4_oz_6JRpXUM#w=%-yTg;a94D{;HxsscYJ3Ms06(_S!iyA)z#DXX^LJ01h<@0{y24y;z`(guDf`b7O0
-5r)VUG)7X%eansLNUgbn#A?|ixug#Ja5(#J-lVEjwzE@+6ko{Q^g=@f8x}$cU#o_1wr$|*6=`%Y5B=X%BL*$3H4hLr%2n>|3BEEAGA3F8a!tYyGJ90%X5}*^eSVmxdN8Ka+(NEfy7Caq3Q>G0)GuYLUBor4ta(uy>;)d7?$Uy-KM#l@>EZ*?@
-80tMA#jAP4hVasWjaQI_q)0GWH`0(`;rz<^v{s7mZj+!t&%N|Yp>wu7rkzye!E1tBUakx(ql&N5*3@HZ7X%?BZeQdI7!pa;z7?k5PRs+W904N?}q4
-^&+9oE8L+@MirpRSMBc@KvyHQ%ClE*VHu)jt>sPpsxc>S2_VupjPfk?9^WLfEo)g`Ex7~gZ)l;jDYVfg&0+xP>fjh~cfPx}ui2
-!V|~l2Bj6Yi;ZsDLEzkM{7V@?!j3=&Kq^J<|mC8YX{#}YAfv&2p%VN!#NW9vIv9EQs{)NU@sL^i%YGX
-Ex}F+8AvuHuokiAU!dYpzUZ2JlF)Xh;5#`d4fRQ|uM{isvQOc~MVL2vGv4%ZOh$c~b}}mbP!eC1Jh!q
-9cEJ$Uc-dX+=nx$dF8u>c@^x`d#)Lq4HwnYfAG5z@aGV>XD;A^@Db&7Hfi&XiZ=kTQ8ET04MC}8Vl?3
-?Vpt-)#f3r&|wxD`U78^LybY@;S9hZI$JfP%a*;52{eegj(QI$k>k*SduZh@rzJuoH;8UYzSPC0ldP-
-Ky1q$-8HqYGG6gO{of65g?7Q(B$Q>ta27WH(+kBH{^IjjTb+g;)yS@`
-8-PR-d2YSvd-H`fs4LNWcA>cAw_n}|}2(dxU2QX|45ya9f|B*qXZhvS#*lOFSw+=?QoYwVaG^a~_%Z#
-%X*2=B6VP*__f3I!xciq)xUQXJ;qKdReH8%~z0^zG}e$?xpysoar#E5s#h$m`+
-P6e&<56#J*L}p#SdN@ihRQdowu!Qn2^J>!bVs5Fz`KLDhwxt;5Izn@tj&%C{KLELN^TamrU={I
-Lw&ZHENEq2bJCNJ_LE5f@ZPMoCkt)4N#BRobzO=@iE7FQ0z2&bR5K^kNTEZW7RGbAU-Rm*p%$ly+t{@
->DqZ5~yS=o)R?!WaCE%-NCE&@XVhi|P@%oZUgQ{B$`(3ghq5oSIS#w(yIZICp9be&4p=_&bx8iw|F6a
-xkWNS@RjhZ_KLd^kZ)6JTDkc$}F-I&l^^kN2V6sPU{aa0TPr$s^ZPRx@Cida3g012Oc8@`3e{hYkEu6
-2^SYOydTtrL5b*pIn@xBP_T-#xuV&4p{u*{Z5aHGcj27E-9q4GvqqF85{fx)kMW$jvugz5e0Hzy9SpN
-pRl;Sbt6obg3erECVw+q3Wx!`xbQ%Y_UqfmqGAW&egg5A^uD>mwNKx9`1O${#3UAd^Y{<$FFA-9X_GI
-hfmsk{{_ich#&FQsCgj-NF$bHQorYpJzOXm*d0
-t7_|AR#Qxl#sEn6yc?c)%r;4M&kZ$;+##_^UyeYQOVKoz?G^zz@+XAOu>$w0--q)ZA`c_^Ky;>dU8OJ
-*6%^^XJ9f3wJW%YLlWhw^jbBuv?V-=&IW~e7z-HWu5@>`mn`nGrA%(uGcOs5gsR`O<%f3@Rj%A@yH&K
-#1V+TK?s04awJ6%eUsf+3-BSOpX#c>PWf2~nc$G7cF8*w_4l&|o0+C&{{@>P@;4$ezZQX;H%R*$?Bn9
-ii(>YiXtF(RSuQ`*YX5;1E)4(VDgZ`}Nmf$68C_55%~|O3!$p{#Cqg_n50fd7x@*iJ?6d=1b!@#H6Xw
-QSE!Lm^%|OZM$j)BPz-&QhfobO%{yAHPn790H6MTdUiHu?~83cxUEZTfUv1_;tWSEH#tFaCmMR4CR^F
-AABxYpQSK{D^*EPRAJB5G8%|vZzLF+c4iu<|``%jI&fMzEowll%;yl_@8|Q;Q+?5_)?H>9fmQLmTOOy9$#nIu@8J)v<)An2fsb}{-z?A{6XeEfwYd
-5AjLOg&%`=%G0>w0|7{of8Kc2HO;MG7e;yU`bQIp4X94oX$eCiid?UwFJ+iX1%To^AESHYjSxQYEN3g
-vB)#o4OG16{4P_?l}}+%zal+H)SdoxnW9QUHMc9#|wT9hnc<{O|H2NIyhy}s??lRr1wIq_JWVh;*1@x
-o^bu2TW}UnK+$BXADpHmfc4VHZ|ZC@KWSC&tvd@>>6%Lx)m5~{q2jU?)&01MgQD%h31WM-kaw6rN^IR
-{d%xrI2WH+5-b8)>+?y!R!+bRJ@tq-_b|vTK(Q0<|X8sjnuzQ`we
-0SV0XE9Q~UTGQ2!d+;pu>ObMP43+h!2c{|^KXUzfqXJ|d=Z**K2u|7my#@7n_}-^hnyF!&!(O9KQH00
-00802@m7R>_yM1-u&o0J&TM02=@R0B~t=FJEbHbY*gGVQepDcw=R7bZKvHb1ras-8|cJ<2IJ>{tAS>T
-P-CLdoq_)%DYaD)2&Xs(igAoPA27eX^Mm-#uTX`NV_#ERc-Az%+~(K{=@!BzGTk@0Kto7b&{7^SJIM5
-;Nalk;M@Vv^Cll6xhj)9Q=(Eb7UiPItN9|YO0f!~yKpZ3qob=uqQo+ft5k|N?=P>!+jm!@EY_mTMY3G
-wMJZ-Qz7%1$E*D8Q7Y_-3irZDP@`EHRWs!yHEi^yMSF#98=?j7h|H%(48I?G4E~Zk03#TxW0r@OUQ!z
-_YsSMn5A&*ow)d)hHcm&TXH4+LiPh*kgrHJK9X0gbr`O-h~Jn!g8V;kk!ESayuLdn8;R>}`$noP};G>
-^hm*1zU+n49^z3d@Dlwy^EgS{)JU2~4}p^HdocMT;=WMq&;WUQO2{=(Cbx$&JhP3JIrMUj1-B37)^2J
-pcCW?KdOwB8ke8I4hBKc`*{N69vED(Yl4{NW9PB%Mk-2lHbG3^TIFWUn{9f<-^*^8jlNpJc2K31uHLm
-dM44r2a2dXHESZC^Usm_!s9Chlf|+CU{zY~0JRK@yJ}@1Nfm{wU8CY6SUSLWShQJ_Ajd}n{;2mNRJaP
-Jeeeo?*KeCdeT@qerbqSr27CHTS%z^~hNk`3p^`6v(nT_@aHhNZ0hRH8c%VvTRo(rJe;5HQFuf!C%jE
-PdnvMMt80E(BlyF
-2@K*y&2I9%^=5?HR$Q5{K3!&Nq)c}^#j5i;W6H@#EBvTW)VznM6ZVY4*!TLPK2Tdz{;}oO5*}6fy@}d
-AOkf#a*5K@f^i&){9YEbG=Dhpl6jUF(i@7;`r6w;XWoz5h9|W3diqntq5o5!WCKLdSF1EB2fz>cLoiL
-l?C#{MsN@h=J4!>P#Cfus6h&SPtVGt+2f1~%kYWmJj*KTHw(?1u>8>x55JpvH5?LVTOAX^R5|o)aicQoBh0jzZfsH~EgqRYcm{wV|5cv#wIf+KtLuMqn0d`rg%@7O^+T~TjC=
-j@%R0@Q*^2H>&4AYlY*
-QLO;84`b9zay+e)6!`1bH!EeBR!^Qb3_qh?3R^QNgP`U`CDYl+{`$23dIt=8ES_q)ckK~Az^&dxSY&{
-Uck>iBj(KdmIU^etbB%i*+M=|m8*1Vh>45v1_U%pGL@$N|somCTjLQYvYEKyS75(>|D@`fY5E)F0R*Ji9wya
-u?s(Fw7GZyy}BMroek$8BS_mI1C+d+b0=QIDPAW{o}E>CQf|!jw!D;jJa=duPwA6A9kQ%)8
-z&oGqAcX?(M>Ft*pvZjQyZ^R#9vc;&Sqc@Atq1to-@m>;SrNrF5}*GNswO}Drc^
-_uhA<}l6E8p~Rja};4lg$BZYwkX!`+t1(aWjOc2hFf)Z2y`$|LO(xu3^s|w3MFMKe6tlctS3ozn0rQZ^sK4P-FYHw3;)9
-0KME&Xge|Jl?=N<_jNvKU7H}g)ZV$Gb~TqkW$+JJ1|lR9emx6UuPXkv8Vo!0JeWv_&(nhZCG(8d;Ihy
-p5zH@~WM`ISw>-`T-G?*pHQB%$3uf!-pu=fYR^URB>!{!f`Y0$?ftlzmXYw8A2NO6+NBE?di
-I#Iymx?5aX&~7pm^;0tMs$*MP0EU4Y*tN!-TGbU5Dk$fV>h<&$8Ptr>F)LTknDGSIx?}|I;7O(sE#?@U
-@K!$&b;Tq=-N$*;Gx)Xq;H_X54SD0HWnuOYTcdZk!Nk9yhyJQx#~VllekSyj1T)iW?h*}!INv3A~%BL
-%NWyU(rHsDi!r>q7Ppq-;QFlnCd^@T2DRnu**E|osZ=FxP{B?CgMh$Vz!2dyzXyYhXqBw|7FlWonF!-
-uN-54K=i*PFe<^053J_*2%Mj)a>SQIOWR^t4wZ7-zH*~fe5$(5SU}rHtj$w7L-;7}~k{+r&`v>CQ9Rx
-#uL%G>y0^b?Twfk?jBzf)do~S8W|))#iXSmx7Yx`DQ5CnhFX8FiynLBgSxmFJy
-(Q%(!+6NTKH-Qq(m5pEo2~^3Cm0EqJFYN*RC3%E7DuPemf;%XGI7Q2WI=LFXDG-6Rjp{^A(6uvCJdHI
-JMH1Dy_qgWBAm#6&z|!f)8lv*7U5FEKW!2=pAt*A+qOnR8IIpDjNG96R1y;qJIJ`fqyr2T1C@3$#L(|
-WmW(1q6(kjYv_n=7UKF@;e!HZGDejE;JQ44aJrFwbgA*$@5wWR^!b8J=L>5uv-_mTGZbnh>M*BGeN<9PzqK{?t(Tz^;>mfK7^HPXwt_amwUCps*Veh05
-k~o+=gVo5(M)zu>v2S;a)v7nDs&_~PA*|qDjd>Eg~&@p=&+cp>*Mj~H^cTENXZexvJ@MJwfH%qEwow~RWySpFo}
-WR#3PJ>r_g47WLVe(s8?V!2;z~5dl|+h`n!NN)I+o|vWL_b1yht%D`}M@-5RODskz^nyPoDJK9@WKIFw6_^=tg+~XI%y1F_?pt>2)=%W-U^$3<~?x;rIdt7GWq
-@7IVMK?dCrC6gaJ$W=uIOmFKrIq7G$M^~CYl<3j*B&`NC1ihzk
-&DWM{(7=9oqa8h(#*XP$F}SU8HZ1091Ud~&_RT<}3>-D}YQZ@??i=3BsDVvqppN5V=Qk4dpt{%aBh``
-8dBd839e?lpg_J4EN>s@Eq@-?-drh)0jXfFy=Y0mV!NL~FVE2J@+A_!nBp10ukQzvT?!(S3ifUC3U4S
-*c8t$R7bF_14ja6qUIrOF5Q9Vw}y5i_28+jjN+gUwJ(-TiO7SQ%7kAmJ+X@YNh*ex=$_6>a}?xYQ>v)
-V-|_fZTbYFQg`wphb)aZ1-tY!xsRZEe*71PB7F>J=gC;6Bn?Y^dELm|dLH?xPwc0_B#L7zqxpPCiArr
-e;Es*=@Lw<*H8afu7+HT`
-+V8tWGA%NK1}?&n<_m%gZ|Jl23PIy5{A{@(`W1bCBwMzV%oo;V-dI$(6su^k%%(Pbyfs*=YkLvqtCs;uGBhrn_9B603ta6!sx!g4
-_7szMNSbPB#XIkbT_S0T+M{W}L&j-oPFoxR?&Nf-%m`3~sKU-Tat&-F)k38e#1LHK6Qj-9X}kk@)eN>
-Vw_!TB8m36omA~5W5tB3C{~Dvl|mk?BR+bruM;lWs{|-K(|*E(U?_fy6%l?A_R~}@*-WJnM|#
-!%cp&eey^|o)KC;2Q3FxH?E3bi+%x(`0bkxn9z!|Kg**xEN^xABLW^~_7W4LQWv;?GX-Uj}g$J;;FEw
-IZh>SY#onKhSTnLHz_^?*Z#RM6bOrBOw}hDQUI;Vw`{tae%
-7cDS#x?szW2C2M_7!Rdkz-RTdzT)K}*>GPC097}Y88^Ik(D3EB+2hGuWDaBouKV;*1o|Egpkcx3Vj_y
-O$TCifr;uO7!QYhwQYUfl65MICx(>18;*(`}*hkzF|bLJG`S%9?|r1=9zizP!Lb^sAqSJi`$Bvtmbg!
-yA|3sj{naYzdaL;?n^vsIFzgwT^w?4B(+qZmtx<;Wynl#cp;_x#er;4@I501|u+w23A74k(9waf&tcc
-L2cXZQi}am7xQn#Pm{@&`kP;RZNDvyqd7c}hVh3kOdXMVzU-2d<-
-ZoeT?bP|Z;Z$HVxmW54%#mnN!5&5?SjshO!YMnpm8ftBJ%*g1q2(<<2-d(f8=uV#RxY*H~IERBebSOq
-5YUl#ZMOyRE8}f-SIDai)b-c?w$-7Bes}(P_8?qyEUHU)QL^AG#XbnG0pP2E*;grjd-1h@{{CnDGym%
-?((cp&noH?s_yrBCWl$lm{+5BMt=4VKw{5Xe|i^|Fdd2m4(_%TpPp$7(-jyh5-|`1h$qy*0l^twKr=H&Hs+T?
-wa{=?Ap^It?$j4d8Nd)7}|XtH+cy2?@XqSDj^>LVT!NS6mbLaa>6npEz#)GSkUyjAm1hKs{NmBvG0oZ)mw1m
-gXT+AuOUrau07DIJN+05yoJi#Id%#O)Y~W>u!z5FFJ`@D$Q}%u8{X6^1|hcw2&_bXEW?<<1G5HVuxNX
-al(u+8u8gdlAj3xC25FuL{;V()MjgdbIjS=+MSI(iX_Al{GnwbE9#r+i+NsxsT5%2T@`IT)isS?SDN%
-jLJwg3t7J;1SDHi^eIJ@NfKAv?r9a`RuX?F^$0RB3j0l|R{5r@|c_LoqF%T(A3b3NV%Hw0xb)vE7q;*
-xI$F0{%AG3aW`rW&Wv#Uw)>h$tet!vC0oHTQTj)rL;i#g+@2Oe`nvc11}-IZKmL>n6l+v-SHij7;)l$
-4pN?dg^kfLp#G3oOhU0fBDe%oF1?=wqb1sEI`qtY?(3661#Vj@3N1Q=`%0<0zB&+TrWhCa&H9F{x!rU4BGzp0o``lE`x
-eDu%WSlR#-j-PJXJ*r99AcIuT@+D-AHQBvnD^c^9nhgAkK(h?^p=&7M_=cM5-4a@f#^bmzU8#RTr#n=
-wwd?ni$1VNj;>{y!6Uw@IIS>29{$-wQrmKr%zT#pK`KpIuDeOx|8yx{LPm?
-DfPKT6faAR;!pCE0;!QX-5J)2i*1C$OTlJ1HO2N;_o5B>IeBDT%xz&Letf*n6@h>&0)IO0zU;o9g)=(rgIRh)$k
-FH0e4d}IPvzIU_(5|d*_;f+UDv-FEguoYL08lXNQ>67llC=b;Knq%kt}=-V49tSNN+C-H(gzDZ>9P#f
-v@6UFGKk^e+9H2r?r<7-OnGnUWFvL@-inoG3c>ieTr6!FN!3NXyqSDJXsHIR=s%w=#9Iw|sk$c*&{^(
-*XTDjU1mnlqbQxybc|h1`a_FWI4Awg=s(N&tsOXIDHGVm4x7lKXLO_BJoSmaoYb8*~q1)@XfSSL
-C9L&$$$x97Ru#E1Vk#GD*;XRE4W^&I;qyk-cnf%7e(*$z0BOfA7CTMyK#>Bll)DpQiX|-D`i9<*+dYp
-E0KzRQ%F>In#xH^x1sHHkhtVJv7p5TqK6O^Ng;x>;jlAhWKY5ZmCV(bL$j6bgOUlg)R=3yNY2uR4wsE0S0YmU&^78ck>#Jb$m#fLe+tb&te+b^3z6JC1MKHOzcy~c#d)|*nXld(
-R;U3api({kK%4cyyWs%D;gJDqBaufi=F{pac`eW8Rf+k5T+)o!*iF0j?q8q9mAc0r%*@zeE65~5y?aJ
-(^tdVoi1JhlH7>EGvZ|qD=NI3ZGqh0!{LA|Rd7VS4YkxAWS@5xIfQdAWMLT-$r98wtSJW|PdFdQJQGE
-7c|nLy54Jlglg7hm265T6Fo9CBayeVC)cSca3pdZlQ4Nug9}<>WVYRkZb%GY!i*JDhw
-qn@VsPLsVwW7?~n$~2J_Fwg}v=u(&AowVL06(GaDqu_5(wOHSN1){&ZY#uw(`3evWJurR%OweA${Cxq
-f2%QnuYExV-QFmzjPl4I)P<#+7dfXmpT~tQ6msTN%+s9)9WKtw1xxfFc`L1KPus1*Q*e!niSz&rF;*?
-9x=4k&0^_mBI%>zDVa3|Mbiz1(!+enG?-LUTK=fRuFO-h3=5()R4L6)QA(h_BHvNm1AJXA
-thX3Sph7I=>y|v%+N8J!YUf^#dpV!Q!GB_+f8jChorOvjW09=p9E7;Af3h%{m8JakF99{c==q8nY>QVBbaL{JY7-WY!y7{r>DMrxVH;o?v^U;=}dn7^1ue!Cw`id>;)dfn_&N1
-Rrbv^VJ>iXnm=RU8AQ*QoeG9X&}lj@jR)z$(POUvpX^wM0)sIMvZ&*9U(x|Ln-ea4C7gtIx$7xC5BIa
-`Qy46_)JX|&2&;GHz1>>Zxe08noa7$hjvjQ#KvuPSHbFI$z^roQMz_r2~>+}=%x74&rqSRQwfVCya_S
-$g%=&_4Qp;MA$Xw{qj)6Rmyt
-?aZnLX>M;y{%&&dHH_8_S^*jX9^=us7sD`#notfkewh~o-KXXIb3=XK=)mJzW|Wa+v;DL34CW&X|HOR
-hxba9#=B)n%P)h>@6aWAK2ml*O_EskVqo?Qq002}0000#L003}la4%nJZggdGZeeUMZDDC{E^v80kil
-xhFbsz8ehQJ(3Y~o!Yzu71oyKN+NJv4cNQhtasYJ1-tE{6v-IUeB+zDmX&JiY^EYaP^Rn&xLBg_?`@_o1&v~&MZO9KQH0000802@m7R%wcA+pq}$0Qnm
-L02%-Q0B~t=FJEbHbY*gGVQepOd2n)XYGq?|E^v9(S>JEmI1YZFze49eWDLwWEpWwU2gn|Fv0Yq%-d?
-aRun$2n(Ae^f8js|Yg1*xoa|=)k!d??jo!0bcb
-%x6)XD})rubeoZs#C!61=PYaejq^n8ggaHpK@wpa#wu||U!Oxwtj&~i$Pw$M;W)WV~68DFDqd&p+e;#BP
-7qnjF-_ly}@RCS5o6Q>0vbD794x+Kv?bG
--^W>{^V!vQuuw*TU6CRmV43y%+d;TQW{h{~Xt$Xfv^Fd)`MYFbgFtu1W;A`!)Ijp=Xa0z~A9%);lYV}
-D=bya7M%M7L<+uWWQ-8MkCqJUT~F|*>S6`921}WPS(7i>dXtZi4w!A(7E;g_pL;zO20A|5&edYohloUBtnz4m@)G_SIbd_^6FGP=48kKIBTWd(HaQ6@
-h!w1pS2%XBZ)QRo(s0Fd60Cbz}l+T~3FEeJApl{wb;9usbT6R!|>?YxDK&6{~yjBCu5^z`7Brsa1J|s;#$GPvRHo}Tuq=i`EHcq=EE-5{;J1#TQjawKi
-b;qBvLSy0ODpQIlo4REm%^*T1!+x^_#$AS-kBkN*oXY(F1uL=!3xV?OBQ^qPUSGbRQ^l;x*}}1gTIJ30km&aEqXdG5!`+avbZOL%c*Q4B(}Rny_M{nbm
-lJX=ei?D1U(^y~5)4d;`D$Pd9W{$~^}qYEc3l2Y{?*6R-hbbr>PsOWX&62%$Y5ir5b
-Y=<8E+P2BNDTCHBbE>xfJM<=JoXtM@wwM(y*y&y}yQlbF_X9;~Jp!_08M2KTPGBR~S~;Sq6T4
-#N;x>%f7(or^Aagb=683RIpIy#k9Gi{A-$YT!~T{v*0H@PV5vgqQNK(DfFcD%b4Vz8gDh$be{13$hd_
-ZcNf4WD|nyzHgsF452+F}31DFJUZ&I_5xesppKmOtH^L@Yhf#Ti(*U%7e8OW+Ou!iqdq
-1Bf)f+WEi_yT1>JEqpxFhxhbOu4=U}DA@-bg>w2C_d*W#Y4_2fs9eW0*Vwh`^iF8}KSY?
-1vp#n1Fj+E$-|a3#h2v^F#)Wsv4kkD>4v?XsVi~cMeou4~f-|z}*t>4V*7kuMJVda`>^fq*Sa|48aus
-0N6Oi|bxJ1mD(#B)#CC=rU(>HJS=AmES#=&lcVIo;XFJQ&dp%61pS9w$q&VX?n?%0CVEz6nD=q)G~9iUqHC+5HSnW*Zu7B(VtS=KjoI-;hqLO{M^66t_3+&
-Y43_b(3q~d7?cvNq_Hq=E@j-pu3>+-1)oX_vbN8;@Dy!A@>~zM&;)$ACScN75kDr{mpzWa&x7J5AxlI
-JHmXPMCpVbVXsW(>E`sjQgU~Lcd>za(+s_$yZr|^{72Az81pkaai$~RB~aAJ2C1}%pF;{hfL~J=JFpk
-7T%apygF4&bD)uvydZNx(lx6{xzz*TERCPDN4E-HF>SZS(b+Jix()Y({l@B`ax_HG7h7>`v4tERz3Ly
-?zI^!TOEdHsX$tOl)?*Jd-EnG6(r##^!8ijvg)#19GraHvAW?CNWi1s$#=hk#NLNPjuBtS9%DdPIVyN
-!b6D(px(iTxt122^?vv0BD)#<%Y@o{c9!0mqr(29(=Cp!i01(jCyvc=XxTrPCR0LZ%@`>ft$-60j*O9
-xEH*jxYiX(F9J5M(`-dPLXajP!F*O=JPCUMRzjV$gM=06aA%DyW`jot=!9wn-EZ#Nz;7-KkvwCB6~Ot
-$(ANSNI-rlr}LVs^3SQpsN|QtJ&0!+5Z4gO9KQH0000802@m7R(R!
-C9=Qqt0Oub703-ka0B~t=FJEbHbY*gGVQepRWo%|&Z*_EJVRU6=Ut?%xV{0yOd8JufkK4Eve)q3n9Sk
-BbJVFsH+QO)}n`S1vL9Kc!;2)!nk2jRgDsIfhv$B?*w?brEKfDpV$
-Wy7z87tkX(>zP7j~)Im15&xKPp-I=TdIBqT2dbS?fA&REJ8cxK8!XzpA5}6^-1}m1$O{N^!+RM}=LoC
--kufUcdRgIxg9Zv@Fw2$(QUxWO~Wo)>_CaEoT;@7ImE5^D37OF3s*zA}qe=jlwC9$WX^pWgil^dy1eD
-*)GmxrDVxH>`#qs>qn06)8?+N{YkE}wC0Z-)%-#GgV%T-k}oRF8?Y3MD(4R%pe#9b(fISPMRoVWFFp=
-u+2=_?uO09^yZX{BP66+^PDu?OV9Rxq^CE3ao$%^jG_u7oP~q=HePvgsfIu2<&g33e>LEV7dUNsi!`0)*A%f;uOR
-RQ8Ml*&F$!o)v2Ee_{Qo~Cjn?&z~N`Ptk4X2KJC#&Ub@%Z(HC^;NS^S!@4%|E6Ys7my{PH20{vK`Ov6
-7s!79xC1}S;OlRryj?<7b@!vxKeGylUy{jS#t6G`Z9U>%QaiG`PskE&-UkM`Ssb)tFvELXIKB3&t@}<
-0?@6sLxh^m=#C#(45qSR=ReTz%E$?;$>yjLA=dN^v1c%pXkj)
-94(l#5#cG#6d-i?Qfh^)3f5^oaVs-!x*KJc7I8b5&4VhN)f&_I%^w?VsjS!8{NfN;%3&=8sG)Wdck#A
-(>&eMmE#g?g+iA^^+x1^wT3mQLto;ql_C;i|AVViY_p0MlY$h3qo)12uYXZ!(d!a
-Xh;#~)$4Q2;=n4%d>+34t-8`EpWY<^o825nu`5;3TV)OAZ2*$eLHt6pc$}lnhqa^Hvw>qvb0V3c6ih&yT3oO;DHNdA(R_*0V7{6UW
-oVN7NijMtddllXXkbNrXoThw(bH9X1N0{a#S?br)2(+$libc#%`;Jm%xTq@#0Sc;75msKJ;EX(9dk7l
-iSzf=GAUAG1F%YVA=&679yx`ns-q$%D_c%_b{`a7>ESAmF*E-}waR#bUO;1AcFBZjdOszP;I?j=QzH-4Y$Rb#FZ>E{_=I!%CPs@@V8yv=&p}l-M@^w04G2?7
-g4aa}ql)C^>b)vM^x*=9J#2!i+CgY(4=G67$ef}SkuO(~i+_-$xR
-*%regnyK^lV1Q_1(5_u{Khc)>?dVso)X4SEQK@5CL}Aal6S1GE{{hiB$8oozbA{qNcVC}
-)pN>+m@$&zMx)XS{8VP0$+BsF`i0eHe*b8W;Oj+jX9irb)S1Yg3$zzB)D8_Ow!q7>C#(oehA;h-zFiN
-3dCM|W>lj^?%d?Kga%mQ*q^voV{>X3iEZ`R^#0LU{6!9MkHK=j)d#Vio23u9H*kmZk~Lc&P0Icd0{Qp
-*&NulWjQ8T2D^0dY(@2Uyk_TH&GI*Xy~G<9IoMJa5(t;{5C7
-+B2|q|3!24e{*Mj}nxJsFzG+8$_|;5%%|yu=}uzEE*g&|6X)(e_D?KgY7|x1Ig1wop2sFXM{6ricX<6
-msCKmrz~tXlA#MAHwVt
-|VrBv+DOMquK&xDsVoq355*}1u0(~56Bi*JYiX=vVec9m#jSbLu|9W-#K6&x{#m|??#jE$!k^_@0>}*owmPfiVcY@*Cb|w;_31Q(I86iV?vLddrl&<(o^~Z)DO;Ca>E|u^@kYZxK|Lc$4=U-&lygX%3
-Kd>7zoFU@l$=x7d}~8sh|6B%`ua=%N5E?K7{+TFDwy5{USfY?0-;80|XQR000O88%p+8=?eQ`WDEcRLM{LR9smFUaA|NaUukZ1WpZv|Y%h0cWo2w
-%Vs&Y3WMy(LaCzMtYmeNv@%#P?Mn+Hx(W!BeN4vlVXdJsRki>!Qq#ulfz>~PU%4sDq;v||uW}
-7IAHk<0ywf(^A_HY0o^V-U+IFHt=gQ(z5+f)Uwt6xOk@DpJ4hVDc!0rMq%dWj{ixdYJhM>U!<{lg`xYJhy8
-HY+&D(*XVhMfeOymX?b5g&-znun~d1sA{38|Dlm%hx0!iyy^P35YlnNy3VKJbx^R-4rw-kZ{huG5LE&3Xi?A#Z0sk8$4HY8Z
-4P5g0C$C52|qU*6tn2r&cv(?T;m%Wi*t$>d7f-Gxi}mE$G**b0c(4a744}5iN<{tzvEZGe0=rSJiGoV
-xtiWS_~;4-q7+xZ{QGyWUO#{Kw>QrL*;gpw*Fui<0Jtcgy~8N?Zf+jeD*i3aTyHj;QtVk?w9UZ5&@7F
-y+%j4TA^b^Vb4|aH<-C(R?@p+wdPQMLVn8_ExgGx)qE!%;3s41ibBd&rb{xznvh47KG
-hPk+J95A%k?$UVV~1HJj1j9xSJCh>F;iA=n#-R1?c2F>J_VJA!AC^~Daq%_$Gq1|vx=+n)d-`
-8q)!z{2h$HvsY>_(7lqoV;|T;;pW@KNoN8c=$J$q|;QJ@9z~q(FvHfdN<9oCwn81LI&ZC((=g%qDPY!
-L@{5@LCFYt4eC_>`It$n6|_CC?n2D`Q8Kx;>~TTFu}y>#G8_}O?}oF;)$j
-m5u0Qax$QuF%69AvikREW{NZAgM_dl?1#PJ<;rr*`v(G>OnAvRPnnt(*?XS|h%)2wh$)6zR*=!Xjo6S
-r%>RTvuJE;Djx>r@x&&_
-3j8!Gy3oR47Au>~!hxIxjL@%mT?}pf?s`Wd7*lMxrvpUt{K-}y+ADbEl6&pp0pnF(JrV;DjLxzS+%cuR4WRiJ^^qx>A
-?1NoEf@Av}jC}Es?$=-u%~2E|8}YsaO9M;7;}|1^%#Qu~LH}FIA|Lty9)uH72Esq_A@=D30V-0b5>2s
-%2YZmh@<~4qIs&d#(tzU$9W_e^G!ghe#u3vSS(c=pj2&2CqVA%5i|cBBr1s5
-9|nhi|hz>2~;haQ{tGA4w9;T#u)i3ykC|(;NR#Qm-8U5bay|862h!1ODjp@mhp&MlG
-cji{dhsmNpVCwo-IbH7};>W~ysqwB661{za$9YBWv@QOiu#+FW3O7eQl8A
-DI%CBdUcZo69
-!oaWK0-OaVVynb#wyJeGF21j!)cgCpcxG`Vo3&uZuJtz^y#Q&*0=CW9ogSv2=%i#eT(EU~3oe3qa^F|
-f`Y|DjKj}2Z1i!z_a*ZB9t$1X^Iv*n%>@BGX;Tl}D0b(a_)Kb@+2G&o;OlGaD-OGjl7suXfC>M=k+wN
-(?lWx#~m2cD(zi4N@Mr3`ZOfl%9P%T+>1SM&`*c53h3b?HRp9OUA42|m%YUT{1vs@9K6^jllY?jRycI
-J@pXEMa^8$!DMMW(A*lE_>!Aq~&=fi%YQ70X>TMWpx0C(q2vPHJ5iuF2sqVJN}7ltPmp|HlZC^U~L-s=>+`O|>?Ox~P_Ba%bs4X8`@W1YEPXPVD
-zC5yg6$BIrzGb&yEK(SKx>-*jX1k0_p)sil_d6_Qt!Ryv<{9}iz)e!oK0I0*1XN9lF28}?4>a05}nzX
-u|f@7h_5qw3L#3}jVEN>s
-pE%7pitq%!p|H;ix3dF&-r~_5uTg=Gg$O5U_ecn;jS2D~I3`$T`i^pr3!
-#1a)_Jv-X+$Aez~r-DN(dtp9;3?V*})`6CM3r!u#r9bV^7H~gx=m0>;SzS)uvIiF|KC+xK-+FsdV+1@
-vG%S#5wy8~!QUsC(mtPWzN$iUD=-+IY7?KcBBjJ)T~;9Qi60sa&mlA$_N-V5uoD@YeyR^9UgB7t1=?R
-Yqx5Qhc&hiveJZ*EP7n6z29`*AC7AgBwWu&C!Yj*x-=yu7krPT|_?OSnEd;4_F6K
-urzQyrZ@PrUTUIou=^~x&g=N-jam%Fnf@MUn3sx?iK*|Ff@-ko^-5sHC-Dh2j6R0W7k(o8*N8M5K1|~
-i^OXy?DRTb7T96!dk)gihyPPC8{{&D=0|XQR000O88%p+8&UflGXaE2Jga7~l9RL6TaA|NaUukZ1WpZ
-v|Y%gPMX)j-2X>MtBUtcb8c_oZ74g(Pjj$fq)2mk;S8UO$z0001RX>c!JX>N37a&BR4FJo+JFJX0bZ)0z5aBO9CX>V>WaCx0r
-O_SR;620qJVClnO1LE8=FVBjt^q)0dVv2S4$yxh%|8(o1sR;=Z(ASP3{zUhoDM!B}3^JU@Z9!W
-=^OJ4i+~D}tF>VR1UqW;dM6{V#UBf{mRLH#(=_#5CPeO}&Bx)5VVMX;hqm`{VO{@$3En9zNWCzDr0WG(-jXHg{IpL*qv>-O0xMt)7JhFxSDlO@c~
-_Kkq;O_9@xss;#+Bv(1~;c8~x8FgcwF{fbh_HZv;$nqkkXlM-MBLA&L~>bWIvtXg)iSgXnVX2;HVQVK
-aIiZP5ME^yW}F_YT82gK318iF#7BE_z9&VUu0?ZUZZkQ^y08hT=@KD0MV%PmM^JeZZ}e`6D%ZfzV)03
-Z3O;AF%6%ub*hMWQ8-VD6X6IlUP^8pevaGQlteu2aymiqo1@RDoGHM^?8tp;ib$kc*UG*J%=G
-bVKDqWr2;a@cxHvC8C3hP_id3It2`B$aT+Ie!FnOcRz31PS0-mrQZ;Ljkfy|rias&g*yTlheaEdwu~
-J36Z1BkbQ<8ha_Qn{&{7yJmgh0ji_uh^%XXjTbSxhFYA7#f5_@7FR*lMWu4I)Kk~0HTS&j}_mQXhYip
-uh56#k_wm-rjR;U@vq;SU-K|3W?)X^UgtkOctrB87}J#B>n`0Lm@Nk6FY{M%i}E3AR)6(yOt`0-Vfhv
-UUp~1xDA5O-^#Nz8S6&MV^nmIzmzX|Bjm*=%8j;1@*nzk|AsaXL70UsJ7F1679
-G&h(c#vQ@!dGl}M@=$5*2=S?pu$N%vzQ;7c|;o~Q4FkY{}O8UKy*hkh+-$p`Ez+C5X{Co!kInr@NGr}
-nwhwqEYJO;ycTi=!2!Odm^J0VH_7j)I><~-a&;VyWU%yN6?`oFNAFrEC%Z0s5~^L$}CJBQ5*`&@Gg!=
-!)Z*t}~l+l6@zh`Pc4S>WH`P^d?(C&k{fQsCjVERjqtmjadAAGpeLDz<74lNCfHJ@w~PSk6IXhT4PP0
-s5&YD!B=bvK;z%N6Mj+@6ln}`B;ZV;xmX&f)P2fI0QV9Gt$fwYA63l$(&%5?L(sh1CyZvhr5NMnqK@9%7cd5CJ?YzG#8x5ER;
-6(pkDP36$Eh9LjUV9=`$kBpWg;&=#C{qHw0J4QJ7ydlHTl-3V3cp~uN5%Fhnc
-NZY=uBN?w2f;c6Hxfvi7|z{*mIWUPefqL`OFow_o-1Dh*}=HCZ5OX?XAbeGJycO^v
-L>G!Cka?49;AJMF}?Kl}olPPSJP3jtu~-F>d!UDu~Drrs<6>^LgpVFH5uDJzWWvSUtdBW$a)X({&^>Jj8
-GVk-oy^5!7ot
-994L{BFdDM35$XLG2bU>A?bD5WrW~3Jx*wXsIZa`j#(GB0itA4x{^N$m?lZg1wCceE(eS0M;hoW3;YN
-nzV^X|uR__Mjg8DCTC0FR_Yr?T_ye436pya;+P)h>@6aWAK2ml*O_Evldu6ol7000&u001EX003}la4
-%nJZggdGZeeUMV{B1i_`Sl3nTDPNXr55@GF>(N7~2kxTs5MXpk5gCH1<
-)GpH|XX_%3a;+0vEhKugQR$}l8b&%@t9n&uc_xgN^?QBX9_pXD5s_Sp=*wtr^sYX`%I!}gDWvs#`5^<
-5YUd(7Oxcx460u6;l-(+uPuX3TE1imD%9cfz$dNyosVwBLGL5wfjf%FcPn!wLXoEN0v`(rtmj+0~C{f
-kXXcuco%boy)E}V)vbdt-rygFEF{>rcZtlsA7DGW{#ra4Hu^#b0zohf_DVgX}quH{-3Np63XN%ldToy
-dXa6k3~$ys+RambOAZ9x*&O4-a<_Q}^@!;pWr)?&*@yZ5DK0^NL$MbneVmv0?nl|
->h0Yd`KokbFAJ5#*UGfEo>Fy;Z%`6MN@BycN+diwm%G|Kwley{Gm&q(-lxN+@Mp`&sSaKm{Y9pad5j>
-u^;@1kvobF-Ud7!>d^=^M3fV^HJTrPTWhwbYqW&u#Pp3~qgp9|k_PHhAp&KYqo5S*u$p^VXoS6d!Ou1
-PteLL1oqE{l}w2V~_8;cEjbG2Z86XUBgc_9*+zE(I*k-};lUZk-!yxAX(IEN)UXEXL3lpYnlj7FqM1l
-O$2UC`F8TN`tpik;+qG8&Csf~-7oJP^E-bLTUKg$(BmYy%}SC#@;?s*w=qKLvD13F!5lt&m0m@j
-1DZ8YsMj=$~hlfnXF+i(09mS_W~r%p?p(TyDZ(sGht_cV+);1&kH@a$RNoitPxK}%^qlkL;~mOPj{mv
-h`sVxce_kVvw?gxt|4oU%-k9GbK2P@znrauUn(CRPkv>;5Hr69UBGpf&SqatcH>F+n22M(1pxP1dgk=
-h1tDrP`lq{$tQZ;PPdV;6k9EX(7pgMEtoqSx_ch&j>+`xwJ@-QB1!^TTHVEETu4kFu>LQ6rFjiI`mTQ
-)8l`Y@d2BkZzk7Cg+@Y2#yc5V1D5}C?fMj*v|er(X?8zqjkXK}Ww%Atq+NO;FBcbImmP1*unfZ=>Ak&
-TcDMN%!en4$n~FB9BANM^_+&
-4De9njjym42i*5BJ8S0nE^^%=Vb#lfG!9cdo9Te@Si}YKhqhrJ}c?h
-*c>0>Afb2cXBY;V!AQ2314#T~rq*slAsfGWg&x`c9TPy38EIvjH`JgciF{5hZiPi`n{jz^%7RpLhcak
-s9+`hwXJ~8925;0EJ|nzdoDxw<8XX#;&}uGpKHs5R+<_74+l>aj^AAp#N-9H0U9R-s!}~UgcX)htO1A>Uh$H>hr`MP0vTQzg
-9?PNe_KhRqeJwGqX?IvBJNH@ptUAWV^zmfkq)yE@%0yx02AJbbeN*SXQ?Z8mnec#>I~oE&`?mK&8(tVEp)*kh#@kZW``OU3xjQhNcQ4jxlGAeF<(mzaIlr}8IMmeHtF?)9XD%gnDZ}k_p`I?kZ+L_;FARv0+iGAea>tqBehm=JF$lM_fmi@
-R%M9|DNT`=<}yj9@YP5Zbc?ZWFz*qwHM(gF3sOwk-02jdZKQ~{KE-7aG$$l``5xSsLk=>b4^cou1zmp
-7Oj914vi+8BI#5F$EWe-M^M|L!V!rsu=L`P$bb0@9cXjjl2$f@8(kThuLiH+)a=8I66t+E~a-T|HY0Q
-Gt*aMed!LYVrNode#&~js{7s&~ysg+o^x~?D~A!Lb&V4GGcpkAagi)-?Oo+q^5Ft>;DP32>EmC^B%<=
-yq&Ia?xiBRyhHA@ndQwQ*zws%p&NOSyFE=>5HQ-ak}lsp7*Y;jnZN*cxTf+g2yBM;;y>cby$(7Y+YgB
-YWF=@?a~vxvIgb~%(73HQPjXO91MFdY>FAVi
-deUA>V>NW$d5)ui51_Hl2Berfb#ubB;Fqj-+tsQmfqHS1p5%P$ni-KB~tdBeIV7Q-~;=uB
-1#s-@1AevMc9IF7rsUQ`Wl4oE+8ZgxaYS$XxBI1qsPn<(rHOmn0~S`4+v1^^_fSQPqeO*r?hN?0d`3e
-q`%JY{G%^Nzaj$j=_%IZ^&u7R~6LRyK6eS+_i`lKjFq)zC~R#BbYdSQ8=%!@1MgkoYLu+ru(2t;CR35
-Q2(euOoV$QGCQtNcLgBUz(v*9ae}4mu|NCi<03sA&mGlsG&og{zea}|Dw-bMvFhk4r&?v*&2k#au|Mv
-F8a`5uT@LOD!r!3}T!_07;Vl~GxePu@faBn~9KO%!)5i-sE;Ls-I?PUaG?uGkGY)R5D}3$kNt)kFoKs&+PDwbN`c8bw=s$y>D24$0-&Fd~*FKT7pUEFu_g
-__X`!mvaNfG>^BI4Aips!SKAR5Ngh^On3%3?j485#N+W2t{l+bOVaA|NaUukZ1WpZv|Y%
-gPMX)j}KWN&bEX>V?GE^v9>J!^B@Mv~w8D`w%kBDqUqD2nY`$5u&7_EI_(79@uvA`svLp
-k%Jg|NZ(kFA$VvDz3XaS6LE)nVy;Mo_2)@7llIxBAAecD(
-+_gS-4=?p&{_0#oQ7xTWFr3L)2s;vU3jqXi#bFX!w)@fDiI#%m!Eq|*UUFa&!ExhJFo57rmMpp%nKce
-N;%W{*?i+G@$^^PWF
-e!oOUF(1ybTQyR@r1?hI49$#epw0ChkRtE4Yeba1IWB-YCxGvCngcWa0IN5E(|n{ZH^72D?g#{rlRMy
-zvWnN~>^8+!6DI7EfM?1go27a7Pn{I$D(GtXA)e(~OpC2dzmu|9WHj^MTBlW2cJ|
-22SsHZs`Qr3EIXOOgb(;L}`g{bO+az^0OV$-^t@m9ZR^?phb-XC6RoW!Lq)k=k2HcPI^n@Qi&^s&U`n
-~CNwmp0OK-(s*a8~PWvxF@oeJ<(Y89nRjOH_!LX}#2TtrwT)@7}&lUL9Y&I=$H6X_eJ8(*@YSDW|}sy
-&iHk5a`ITcMRjdp=X1HNSh?XUatotNmXf9>xB1AqLCH@Chw%hqTFO+xp`IJ(3Tsj@ptNNS?Ebm!9U#C
-`1>5;g6NQbByEs&X$ZBUzj>gTM;*pJ>Udn;)C@`~GZO&gPez304E8>^Gjp@;c3RMEQE2e^xX0pPzR&>
-hwTM12tdHDqqCUWjkA4%gJ>vz)wn3z#_%{$r19%zN>pW`)!{M;!CNHal*9POO^wuJrw8nsbPH6vL0Y*
-o-v9CvV^}8F4aPEnd7V0&~#lj(jny7O6o1Qh)fW}j5JXRpfH!F~@4N37iOn~)CADN_2lD{vqysvKcw)
-R#={R=1jWbgYk@mC3`^QI$AHsZNlML1q`&4`2TYPbd@y)=+nMAwElLg;Zip9@k4@=M52F*f|hLWtQ?&
-u)`B=&hvEtUO%zdLlx3B<*->PZaJjSPLnaUEco@6$Q>57ctkcSyx#>A{)lr)D795w2gWgqAo*zX`Gy4wf}>h9O+G~4qAL|v}YMj*}L+Fa9LL&n}TE7EG4ltsQpLiu8T0LvH*N9sc~0gO
-+gkIW-AEqAWbc+z0H={gaD0DwWSJ&Tq;`NL5#ByJi+xt;S(QC2#D?$7p_dX2IQl(-m%z&GW6k(cSb7C
-DKtQywvDwE|Y6LBM2!v9&;|%S|=Y)V|8VwzSqMJ{i|*J{%inV(HNXXA*k_~
-`EqRd5_W-shJORgjS%(Yaf`*;fOJONYCJB=F)U9=0a${S3k2kXF>}Nbv6*etf@P3Jz_^2>EQ3@70O<{
-ZBxymJ6exg(v3o3>l`TSF?S4cH{>iw@9QjEGtdG58glo`U(Zf2teTDa(n^`
-YHr>4ZZgxqq8?mN_w@(1q*%V{(*mISmasoytmZ(I+<7x=qiq5=mzL8^PM*t#m$=?%sl%1u*ebBaBHg$j08y_tfo1Y%eFGLZJ7PUosz80Uh*6)2bLW)8-#$e|(OC96
-x=_ImyV)EE!}b^^G+1A@*s1;s=%GMf!_h0sTe6gc@V#a&jF1;r@t(kjEHuzp3s;A3eBc7yTc_IStdQr&0_I6g*z!EVBh?$o4VH?@wPXQC
-mFnneZJS#Uv(_+Z4{8hZn?m%7v6;Ob!H%)8<0I)IV^Q?jQ9KpnpDA-Ww>ZlD-=7nmqm9AxNMF|FdjxP
-TpI0plcdIxJVAP?PY(ZyyB_y>(oU@+p$i?r0Jvjdsrmw*~|*5a4&*(cbYl%ez>RO9BBIfA@FZ}fl}{=
-?0XfCTgNpFEaVJi$d=L)PX2T;oYGW*{|~l@*X_T^4iFho$=4@%h`=Z(oko>G}D)a|U&C{`&Iu$?+SbH
-+?v{z5|bJ7)Uc$D>951i)%FykP}cOL*^a8(<9ey4Pv`==o|ST#)~Z1{u3~-y+(`{_j#q86g2#
-in9d%Cya&AGvlQiLi8_x(nBqei`y){0r1fw==6%Ujs}=zQWEnVHJ7X@x#J_|qE>O9O&PhDbEe-{Nj-WV0nO*bL0^tR}-5UF0jXijlI)LQ1is>_K$&L5hg9XO>ZCg^p2<@skuKi@Cu5F6cw9PIe{mI1oWF&^^>V;&)LA!1%~FuwP;L;
-Z!qrY7lvKko6z6n~^fviQil@}3#V%NoIhYe}**5fQ+GLVnH5PCFJ3d&83qgL(+o{A@W=gKS7qXZY)Wn
-62^mbO^oFTj8xT|WhbHG;KgB_jOrC8fC*l5(V)1St=n1K%t#?!Gvuoy7JXvK7c
-p^VsH{Og4hO#vd*xNsEUHlzufUudoR7Z-2Ti3j!VRN^*93roR5_h1Dt~fN5%er{lBNLyE8Sn
-GM(n6(}cslcyPZj|PBfK>b4%Y>?=CLi9{+UuC?WAvTHSI|A`XQ3#@`c3@5?GW9_Oiyl)lr5%5rnFR9WL=Z(ahP
-5lPGRPUF!Y3fhW}JYQX2o?Wz2<{edk2F7OLc5y&oX*U@n%0G@r7y-M
-H?vwl;KfJbrX;9Lvjmr9HlcNhX#6AaaRfsna$`0cE5xxxOB+)~0DyPMkKk)B`E)9weHO*(BN{6U25Oy
-`Vym&g2P|p3iLngPblnZ9Yfc~@J^WLI~caY{;;?ym@y_iC?AhJ5{02nqY7~topkhQ8RN^zd7)$x$|sORp@cS(xrRfSyu@zKg>8
-|ZtyY`Llz{KIxe<0De)!d_NZ3Y8s~$?CgD%S!i_Lj2UoRuA8b(0Sz`Wq+BWF5_5npVP^lbsBA}D^d3y
-x;ezDATwLxkn`u!~5z*?MuAqY|CiQVy?N+TD4!qwM)S(W$sTg`)8ldbDK9Q
->we+6Cz}+j%S$nQ!7bfloVKO?ett>>yz+PWm*i(JmGsMSbmw0FU3gk@k^ispbZc4<@U&%GkEb8czr-~rAQ(i6ur-<-vb(
-}$Z7!JG%bgyWPcy2`vOn-JP`Zcbfd=MIJU?vR=@9G&q}!qYcNl;|?%@=5FXfz({H+=0yI9E?iDeBBH!
-jzh!?V?stq;t__u2S!*H2!r-2h9Z~j?}0r!Qq
-Nl}E!4fHjuI1`;SmN<4Cu-P{U<Gj1q>yZY+dzK~(qzl7O84=V%A%CD#abQJ6M>iOjQnkU*+Ti3_u2
-du3_E*St1(K?2-74R=Qildc@0gOVQy&oz+*n>h+AwK48B}K~FjW@b7_%j_t4i%yX)_rT=eE|dC@a-!a
-kmDY7u<*EIvz`UVHW!dJ7MU4$GacN5)&gO$`4rs3LdL87)iBM7}IIF7DQAttb06
-kWZhS#YW$m{QZNqyj=0!T_|f@7twD!0#x~uD6d3al?Mv5bX0gUiayz7&Yx9MyaG@Ml@qG(F9AUQWZSt?r`|WehPtn0gJ_2ph8;5ld)Thn%9a|moBUhddXL!01~$4hF!8k%3;j_;i-SCWYjM$<@(N+jwFZ99GZ2xbs@W;^?~*6!R->F4eP;dMV0Thu_
->s&zGYFEHh!7Bbz@8zqO(zGDLl|JM8ZQYktX=wyAB~(S-S$C{Tc5P4PH}t*TzxM!ZEX5rC11^#pl4bz
-A50E?x|(R71%~$wR(sK+o-E>J+wT7^?*KEo6Jg)`0;H=e3<8_>GIxpJM7L6bS$>4~|Q(Q%Y_rAPeE96PhQ`6)R5T-dwjc>rfq1|^
-G9u#1c@-+}|9V@fzoSTgMe(u^z3F+T_#Z4jF?M!l#3%)p=m35o^U(ZIx1ufRnH?i&f9#!g8LMnACNLB
-u#fmEnCkST2jYZHg(40;Q`Abi_fDv)5~mi*+k
-vI71}4LBUNKsm&ri=vE%Kv9Z=+D&HfwBZs7fd$q-*Ck=uDb;$+0bqH9h8pGh)hXKrIp*4Xp3_t-ih
-vKIgWg+)2z&;Q*8J*9F)*u=U+5s8DHvroY+Um?Q_KYV8p6hZCY9nl`@;O%fic|FKE(_I`9rU5r!KmVq
-+)V9F}uadQ1L&N`Fiw)*6=6M`$vIWc8W3gS3Wz`;XpLsR%%qjCw0Cn1V9>^U-FEH_pGBd{8;bxmruyp
--Gx7mjH(eG2Lq=>{g0oNd51|j<`dW-=lVeu$BmR5KwC5v2?RT~AK?^%^on95e-0dkN{MJ4UobS}XH5<
-*A8IFaJ3NVslC$WDPQ3{$T0{+EjGoTW)SSPMIYPYiE|^~SVdcE>ga1kPqP?1%pj4f8+4Xdji}8fLxE>
-ZJ)$3&zoevv=TmsiGL6kq9=Oj)`y-!1$K$yW(3)pl{X)3mB6~{M;ioh#Xuh6-2P^+z+`1U=PBpXk#jH
-K*>E2Q0Cyv)MnTrhHZWePgNGtYcn;qF6){ic7lZV7K~awOI>ZDQjXj-m@65F?qn$Z
-=TWG(W2rW`i#yOMJJ)B@($H=~NazGxdmRbQ;3SY8L{)Px_wSDKceWuc@i+DEvoa>u(wgr*6Oz$
-#)GIZw9=Pf!E8jo7{BP?&%_d(2r_a*OWVwNI_c#!p%IB~tzn{df
-(Mm@d{Nz2BDwa?1bHR>&~AmKZ{RXfb;h?ypg*6AUqT7b>ZdyjG$0j8cX5!Z_LJ@_;~m;t~5-MjCFKksf3BGZ>hYC)%RNi$((~ncxeG;$
-f03b;}*%LdJzO0E_;@Xi{5f_7LHUSA^hiLD`
-8sEO{=$~<$|mheL*fjDztiCk02`KFm&I^Nz$$tpM$`BNc(r3%5J#k$gUO}+`;VsKw{1yEMHy3q-Kt-%
-+{k$cJ6HR5lB1I00|G`}e)^9;Neya$R!pj1KQGd^(0&8;zX`vMC$D(s>=@Fh0@AyJBV+{z>keGdNSh8
-<%tcaEJq0anF70!tkrlP>?v|ahO0}dk!g5`v3JjTvP^)mg-*
-Z($YT}7QvXLAkwZj=6m#?IWcW+mgcq}0yf%PtD_1pbSy(AL@*h6BX>=}9crqcD!#%<-
-Z!no*(4)PKE<(qk9%aJ}n7UWjLu2_ep>d;N0D9ZBnk3;k|@0FnKGdZC3lbE)eCwcX=!ppGC}xEchHH0R(c{{h3JbVx`b!Tr(AS@Ajcr^@&mrsKmk
-c-OF9U%xid5;V}-c;!ee#HvAs!)afwqG$f_*3%+0v6m!EVcZ-Acg;@wWnJzm=h8O;v16|9u#?3IzL^d
-?0laClQNROz)&*%mP6r?I*WK8;vM^<$*gHA
-28$3<+gik$J(Fy*9wKs-}s`Y2CDaNZ)eck6uU!m!=
-!=xUhP<`VO5uRGeUdQCZ!F+~C$$Hc0n_t+rRdR^0a^kb(!DU+hctPXkf7C=GKWsbDC+sj4=HA!N`>M`CnMU-v$|cUq!Dc-?W%?e%%*
-32(m#?uuW7XS;&H8n^a<=Hxuvsm*|Ty{8UZz-4X6t@Qn%X@$T0Vcqfx2Biv`X-#B;>=Ec&r-0ovudc{
-2EQeyX{eZg;VCi;HsPm*q2br0muQTHIY*3)a}-6Z9n;+=%vCwGye_b^NulKIn956%0c8DS^9x;=Q(tp
-Hq;mYfS;?CvAW*dNVxijQQl@V3H7u$jHJ&qsW4onMRG@YFqV_qcg8{vS|F0|XQR000O88%p+8B1cUvD
-**ri1_J;9Bme*aaA|NaUukZ1WpZv|Y%gPMX)j}MZEaz0WM5-%ZggdMbS`jtg;K#zgfI}j_bZw`L4th1
-CTmQL8V`Drc-d^43d3$=X~{6G@b|Wh6hY$7p`kNx9`C)hL|gIP8iJ0r`rxXnEfRm|6uptmKa+7-Nh6P
-dtn>@i*UB76@cn>z84wA70s|Hy>NL^c0(j7mxoT^8u8(5$gkPwHXyCCltsf_Y|?@
-bazAbT%1;nR7TlsoI-!ZeoFm9ncwe2f7^*iYQPz7(BS+|5D0dIh=&-BKB;^@8jFC_@_H`Z$XtJSS13J
-*xt6yLdi(ksU6fiR@1{S8_q17R`#sbS`8mA!grDO9KQH0000802@m7R%ULjxDo{b000XB02u%P0B~t=
-FJEbHbY*gGVQepBY-ulTVQFqIaCv=JZExc?4F2w4p}83542g^T8XyA>cQ>p+(K~cE>_br)icQ;E70Hm
-~q}j0lzK4>XG`-#ew9pGlkq;k|&hz|>Q#;5ksk0<9sG+RzE;*c)wN|AZgRSju&=g&plwX|E_cV7>nQ
-D;ElbhAzu7cuDn&wTtbmeD-+K~+Zt3l5-#TZUHU)1o)IqTk%9)r>+!G;D$3GyssIN|7K#nJuC|}#r$&MPaJ+hlW`Di8J
-kDM>X*PaAFyo`yD=<{n^>g8TpCu|Edp@-RaGXpzg7hTEF2B0)Y(~J!id
-#MB9=B3E3;LhcgUs$-&>iho({d*iyLBFuS6ZGONManxGE5qw(XwW)WOO$avWjhotx(hDgh;b&>^?K_(
-Ej7LvW`Lm2$}`kE6r90yq{drjurM;EVwwX3>OHFSM7GQ=y$tvJ)GBj98wIwyA}41Mh-qP6w7<4~jq47
-lAxl~F=weWFcZIp5n+H{xL3lV}01A?aeC030XSO>)wasG$Wb?p#c~N=)F@D=tLfezOMxl1p(Q@nq~FS
-*yu$x~B&Y3U9$F{69z&Q>e_|J`p1f8)uA5anlOK3t-^~pfL$&S#vSKHnC%KKk=Z-EHX&Nic_Vv^?*F1JlyRk3
-wd`6AlCvw>2&Vl**J^sExhGwEfE4l`H864+`5m*qr)=r7a|44W>+d(WkqDo}83@Kpn7&2csvITfu=#S
-0^>yi9IPgk}9LaXOXvBQZdO<2F+hu+AE46CD|c^{LxGaPxog!96|#u{&#MAy&UYIjl3rNVO3-c91XAA
-^+m(<^=b{yY*uhFR#S!Ad_P#A}mi`g$+`GP+m6Lc_d9r5!)3P1!r7}loDo6`SYa>?q?PiG%4~U#=4!<
-kqM%ReSQxr0dXc^WQ-xD
-JbVC+zNs8swE5c=}l4DGo3&&X8|1wdHnI{EA@!#sfy~_meJH&bQwR<*XDxgsO(Hqi=DVUh@>vs+bOLK
-UD8;Z88N!(+izz3L!syIwGky~mCz@6aWA
-K2ml*O_Etc+9(rX4008n3001BW003}la4%nJZggdGZeeUMV{BqHObI+ZkDEgIXE)
-EqVO+|A?N~uULg*K|q4G0Jo1z}H(L@i?O7^x{2q@hYPHH)HXHsf_8l_665C*AHFC3B{Sm7|`OO4h_28
-?@t=KXYvo@})6c3R)%Prfn)V>ognQq!|~6DG;UmmRCi_#Dm{@U?x+o5OiO$40NY9*)PR{uv$C!se7bkSy=Mi_;of)!&1+s!m3h&W8TOR
-3&iBDPtNlqSIeBnvc%{>t>pTn>=wuH}X|U8C4oN!9W?4q7D$B^4Y@=K;YFLIT&boj^35o2JcQM+{X0w
-8o#2D47KKBtvtgShmGUB8+3crk1dw1(3UmE6-RI0$`N`*3^xMR2RM;M&62=1({9j
-};kJ2OTfM+ovvh5QUR;dcAjeD;v=R6HL*Hc+jI1*6^|#_7CNd*7Tma7-R<}+a?Z#&P*proF60KF^r?#
-{*>M&Ev-iw>mjP-UdTrs8nPHzlN;4C+yz9zR)imfRLooI5a-K?RKOlcCWx4V>0EK^hgA;VEwG~<`JQZ
-t>_*R+t<{e|zmR8V55ix{RJZV=jxSM&1e@A~go*^b9o2Kn4XZI@xMw#?@5mQ`XO#XUxrl>x7p$%(Jh^
-sd#tSA4FOOMsGN((_EZg@xV9MeU-r@y9C^q~yT#>L6Q~zPxJd^+ZlLTaDY
-#;}k?NQbsv#iE3m&*KEuac`m@6r2@C<9&8K|VZx6#7m0u|VLX&r@wS>77NdUq%
-Z2d1a1D;V*60r3bMZqbvM$^T{*;hw8(-}Na7NeP{^FwW=o>8Ds3X}+b7wP4Av6+>hH}DZ4Pd(m
-T%@CoXb=scJWGmDrOT_LpU9KDFcoh>-IW$q%X8H*;WM+(g0p
-WM<@a82vAhcOU{ZAv>PdygY2kZ#7JZ-+v+VWEl_Qv$Jb;KIKcKyKhCXki<
-ERx+Rm3t!)qp6v;JwGkU57*4E)2#*|R6FhaBIK#i7~lyD;-uCnhp>giTl}hS%V;RRO6J(_Ui(j;r`-v
-<7XDl!Y*oRq0DT>M;YcMgnv4Xe^L04r>`u)B6B*`8Jvj6oxn|HHNLkT?FTwl{26<9Qrp4SqLHguzqgh
-o#kEJ$+fengRFR>sQwwPw1-vJQYxA3#WuKQsKO-ER{ED*x)`(>k!<8RD
-zir1Vr1%$Y<|MZJ
-7_BMjTtMvgZy_~lK=j`4tiJraoOEe$K7*mW#jNt}WQra|F*yX&Y$2}uQ*^-)382h4Uy93Mw=S~1LB1_
-`tU9h(pK2)ITV$^JdedNGitvLx}A3>|rMR2>?jh3O#YWcuUgWzToznb*=vws0lO9KQH0000802@m7R+
-NCe|H}*j06{7M02=@R0B~t=FJEbHbY*gGVQepBY-ulWVRCb2axQRr&06bk+_(|{?!SVNQCP~!O438WE
-G|LQyc#q}g5c7k*jos+#NFjtE9ykb&erh%-Wfh5QlcFv=!Xh{qX#)0&ht0JO_JnqR5Btu#YIax+Dq0G
-$<)v#Npf<+51r^0X~b^F+nsxa(R!qNmis~RW_&tIcdOV}b=}kMWZR1asYHt}8h%0C(4p&kCZ+kVjvXx
-GzP)IVIeEn^m6NxuWj$4*&&hi(;l-~**RT^atK(f!a-iI|v?;2F7gA9*$g&bOlaAV#5C8b^@n0Xx*Pl
-Os{2ZMq+AZG=unY>q&U)#;%)`gmeJ}bLC6t9M4qR4_OwVdssge$AU*d#v=$o8K)hlr!M00Y2@&&m+bA
-m2*Oug6d(z0ynft6);a&l5NR7&!xXG$Ia&YJFx=nquEvZ>QZ@vf~IFj5Dfv(*WK3pDzIbU2%{{&53xC
-s(Y$5TO4(3@2MR9`8ma7upbTe$BR|gne>VmZ@aTHYbm0zmqGYt4EJ6ugAxXoIfWYM9a+5c%7P6Iw851
-@nMjPTrff#zH3=cuDRNe^b5T<@E{A&V%2kPpd&4_(mFZNuO(6~9o1w5QvCVJVj$OY0peB94sfT(JzWg
-D=C#`EZ^ouk5hY0Ez(KR6I$e$hGt6St
-aLW%QHV&QBYFLYdWJaEorx7~5g*2hkr_KVh*uE6KWU{=X%>Ygl5`i$rjEZm|IJhEQl2#}kGJ>H_4bR+
-Y_rioIH@|4aHG1Kld$xKlj5P)2EvFY9)oidvPGssZ1gII`rkW-2=LD!`*<4ycd?6u^BP%EG%
-^lJ@1s|YSu;rm?G;rxfLjbEeEvhlT?s1=(x;Sl|0YLh)}d5
-vi}z;n`NGC#ax~BuOvD)0JtV(ve%^Rzg!jo`D2YvtzQyA9
-k-TiUiI7vscy`aX^WDI11S)|%bie!YWtFc~usbE+8$;ZX-JM6UXVx0+lbyM`-cl$%9}$H1e5{77+v*<
-ZREaDKac#kO>4lneoP1cXm0`z0V-C<`K>dXSnkQ%IOZ@JUAk3v7PfLlA|311=8x!>C(Z1|TQ{2}ao04
-qOFO0(?E>fS)0WxqytXP<4P4u7L_@w(nV`AiSeM_7>?Q(!w&sEc&96!qzARYZRi38mLjQB6GE&beo$r
-Xc#Fif8PsZ?KA@rP?JIG2Yd74q>m1wNP>|;9JAPNj5?M
-ZsYh*EO&2+6&^Xf;95eX`px?@Hwkn3$8H_?i^5ygUWRb2sk%m-s
![\"icon1\"](\"https://github.githubassets.com/pinned-octocat.svg\")
\n",
- " Trained models provided in this notebook are from UVR-GUI.\n",
- "![\"icon2\"](\"https://github.com/Anjok07/ultimatevocalremovergui/raw/master/gui_data/img/GUI-Icon.ico\"){kind=icon}
\n",
- " OFFICIAL UVR GITHUB PAGE: here.\n",
- "![\"icon3\"](\"https://avatars.githubusercontent.com/u/24620594\")
\n",
- " OFFICIAL CLI Version: here.\n",
- "![\"icon4\"](\"https://icons.getbootstrap.com/assets/icons/discord.svg\"){kind=icon}
\n",
- " Join our Discord server!\n",
- "
