Synthalingua

Wiki

Read the wiki here

About

Synthalingua is an advanced, self-hosted tool that leverages the power of artificial intelligence to translate audio from various languages into English in near real time, offering the possibility of multilingual outputs. This innovative solution utilizes both GPU and CPU resources to handle the input transcription and translation, ensuring optimized performance. Although it is currently in beta and not perfect, Synthalingua is actively being developed and will receive regular updates to further enhance its capabilities.

Developed Proudly in PyCharm IDE from JetBrains

JetBrains kindly approved me for an OSS licenses for their software for use of this project. This will grealty improve my production rate.

Learn about it here: https://jb.gg/OpenSourceSupport

You can grab the GUI from the Releases section on github.

Grab the portable version on itch!

Badges

Readme will update as time goes. This is a work in progress.

Table of Contents

Things to know/Disclaimers/Warnings/etc

• This tool is not perfect. It's still in beta and is a work in progress. It will be updated in a reasonable amount of time. Example: The tool might occasionally provide inaccurate translations or encounter bugs that are being actively worked on by the developers.
• Translations are more accurate when the speaker speaks clearly and slowly. If the speaker is fast or unclear, the translation will be less accurate, though it will still provide some level of translation. Example: If the speaker speaks slowly and enunciates clearly, the tool is likely to provide more accurate translations compared to when the speaker speaks quickly or mumbles.
• The tool is not intended for professional use. It's meant for fun, language learning, and enjoying content at a reasonable pace. You may need to try to understand the content on your own before using this tool. Example: This tool can be used for casual conversations, language practice with friends, or enjoying audio content in different languages.
• You agree not to use the tool to produce or spread misinformation or hate speech. If there is a discrepancy between the tool's output and the speaker's words, you must conduct your own research to determine the truth. Example: If the tool translates a statement into something false or misleading, it is your responsibility to verify the accuracy of the information before sharing it. Avoid using the tool to spread false information or engage in hate speech.
• You assume your own risk and liability. The repository owner will not be held responsible for any damages caused by the tool. You are responsible for your own actions and cannot hold the repository owner accountable if you encounter issues or face consequences due to your usage of the tool. Example: If the tool encounters technical issues, fails to provide accurate translations, or if you face any negative consequences resulting from its usage, the repository owner cannot be held liable.
• The tool is not meant to replace human translators. It is designed for fun, language learning, and enjoying content at a reasonable pace. You may need to make an effort to understand the content on your own before using this tool. Example: When dealing with complex or highly specialized content, it is advisable to consult professional human translators for accurate translations.
• Your hardware can affect the tool's performance. A weak CPU or GPU may hinder its functionality. However, a weak internet connection or microphone will not significantly impact the tool. Example: If you have a powerful computer with a fast processor, the tool is likely to perform better and provide translations more efficiently compared to using it on a slower or older system.
• This is a tool, not a service. You are responsible for your own actions and cannot hold the repository owner accountable if the tool violates terms of service or end-user license agreements, or if you encounter any issues while using the tool. Example: If you use the tool in a way that violates the terms of service or policies of the platform you're using it with, the repository owner cannot be held responsible for any resulting consequences.

TODO

Contributors

Guidelines

@DaniruKun - https://watsonindustries.live

@Expletive - https://evitelpxe.neocities.org

@Adenser

System Requirements

GUI Portable Version (not the CLI portable)

• Minimum supported Windows Version is now Windows 10.0.17763
  • Windows 7 is no longer supported due to the change of .NET builds.
  • You may download the source code and change to Windows 7, but it's not suggested to keep using Windows 7.

You can find full list of supported Nvida GPUs here:

• Official Nvidia List
• Simple List

Note:

• Nvidia GPU support on Linux and Windows
• Nvidia GPU is suggested but not required.
• AMD GPUs are supported on linux, not Windows, but will try to be supported soon.

The tool will work on any system that meets the minimum requirements. The tool will work better on systems that meet the recommended requirements. The tool will work best on systems that meet the best performance requirements. You can mix and match the requirements to get the best performance. For example, you can have a CPU that meets the best performance requirements and a GPU that meets the moderate requirements. The tool will work best on systems that meet the best performance requirements.

A microphone is optional. You can use the --stream flag to stream audio from a HLS stream. See Examples for more information.

You'll need some sort of software input source (or hardware source). See issue #63 for additional information.

Installation

1. Download and install Python 3.10.9.
   • Make sure to check the box that says "Add Python to PATH" when installing. If you don't check the box, you will have to manually add Python to your PATH. You can check this guide: How to add Python to PATH.
   • You can choose any python version that is 3.10.9 up to the latest version. The tool will not work on any python version that is 3.11 or higher. Must be 3.10.9+ not 3.11.x.
   • Make sure to grab the x64 bit version! This program is not compatible with x86. (32bit)
2. Download and install Git.
   • Using default settings is fine.
3. Download and install FFMPEG
   • Instructions: https://github.com/cyberofficial/Synthalingua/issues/2#issuecomment-1491098222
4. Download and install CUDA [Optional, but needs to be installed if using GPU]
   • https://developer.nvidia.com/cuda-downloads
5. Run setup script
   • On Windows: setup.bat
   • On Linux: setup.bash
     • Please ensure you have gcc installed and portaudio19-dev installed (or portaudio-devel for some machines`)
   • If you get an error saying "Setup.bat is not recognized as an internal or external command, operable program or batch file.", houston we have a problem. This will require you to fix your operating system.
6. Run the newly created batch file/bash script. You can edit that file to change the settings.
   • If you get an error saying it is "not recognized as an internal or external command, operable program or batch file.", make sure you have installed and added to your PATH, and make sure you have git installed. If you have python and git installed and added to your PATH, then create a new issue on the repo and I will try to help you fix the issue.

Usage

Things to note!

• When crafting your command line arguments, you need to make sure you adjust the energy threshold to your liking. The default is 100, but you can adjust it to your liking. The higher the number, the harder it is to trigger the audio detection. The lower the number, the easier it is to trigger the audio detection. I recommend you start with 100 and adjust it from there. I seen best results with 250-500.
• When using the discord webhook make sure the url is in quotes. Example: --discord_webhook "https://discord.com/api/webhooks/1234567890/1234567890"
• An active internet connection is required for initial usage. Over time you'll no longer need an internet connection. Changing RAM size will download certain models, once downloaded you'll no longer need internet.
• The fine tuned model will automatically be downloaded from OneDrive via Direct Public link. In the event of failure [ ⚠️ Finetune Model download is Disabled, Model is being retrained. ]
• When using more than one streaming option you may experience issues. This adds more jobs to the audio queue.

Word Block List

With the flag --ignorelist you can now load a list of phrases or words to ignore in the api output and subtitle window. This list is already filled with common phrases the AI will think it heard. You can adjust this list as youu please or add more words or phrases to it.

Cookies

Some streams may require cookies set, you'll need to save cookies as netscape format into the cookies folder as a .txt file. If a folder doesn't exist, create it. You can save cookies using this https://cookie-editor.com/ or any other cookie editor, but it must be in netscape format.

Example usage --cookies twitchacc1 DO NOT include the .txt file extension.

What ever you named the text file in the cookies folder, you'll need to use that name as the argument.

Web Server

With the command flag --port 4000, you can use query parameters like ?showoriginal, ?showtranslation, and ?showtranscription to show specific elements. If any other query parameter is used or no query parameters are specified, all elements will be shown by default. You can choose another number other than 4000 if you want. You can mix the query parameters to show specific elements, leave blank to show all elements.

For example:

• http://localhost:4000?showoriginal will show the original detected text.
• http://localhost:4000?showtranslation will show the translated text.
• http://localhost:4000?showtranscription will show the transcribed text.
• http://localhost:4000/?showoriginal&showtranscription will show the original and transcribed text.
• http://localhost:4000 or http://localhost:4000?otherparam=value will show all elements by default.

Examples

Please note, make sure you edit the livetranslation.bat/livetranslation.bash file to change the settings. If you do not, it will use the default settings.

This will create captions, with the 12gb option and save to downloads.

PLEASE NOTE, CAPTIONS WILL ONLY BE IN ENGLISH (Model limitation) THOUGH YOU CAN ALWAYS USE OTHER PROGRAMS TO TRANSLATE INTO OTHER LANGUAGES

python transcribe_audio.py --ram 12gb --makecaptions --file_input="C:\Users\username\Downloads\430796208_935901281333537_8407224487814569343_n.mp4" --file_output="C:\Users\username\Downloads" --file_output_name="430796208_935901281333537_8407224487814569343_n" --language Japanese --device cuda

You have a 12gb GPU and want to stream the audio from a live stream https://www.twitch.tv/somestreamerhere and want to translate it to English. You can run the following command:

python transcribe_audio.py --ram 12gb --stream_translate --stream_language Japanese --stream https://www.twitch.tv/somestreamerhere

Stream Sources from YouTube and Twitch are supported. You can also use any other stream source that supports HLS/m3u8.

You have a GPU with 6GB of memory and you want to use the Japanese model. You also want to translate the transcription to English. You also want to send the transcription to a Discord channel. You also want to set the energy threshold to 300. You can run the following command:

python transcribe_audio.py --ram 6gb --translate --language ja --discord_webhook "https://discord.com/api/webhooks/1234567890/1234567890" --energy_threshold 300

When choosing ram, you can only choose 1gb, 2gb, 4gb, 6gb, 12gb. There are no in-betweens.

You have a 12gb GPU and you want to translate to Spanish from English, you can run the following command:

python transcribe_audio.py --ram 12gb --transcribe --target_language Spanish --language en

Lets say you have multiple audio devices and you want to use the one that is not the default. You can run the following command: python transcribe_audio.py --list_microphones This command will list all audio devices and their index. You can then use the index to set the default audio device. For example, if you want to use the second audio device, you can run the following command: python transcribe_audio.py --set_microphone "Realtek Audio (2- High Definiti" to set the device to listen to. *Please note the quotes around the device name. This is required to prevent errors. Some names may be cut off, copy exactly what is in the quotes of the listed devices.

Example lets say I have these devices:

Microphone with name "Microsoft Sound Mapper - Input" found, the device index is 1
Microphone with name "VoiceMeeter VAIO3 Output (VB-Au" found, the device index is 2
Microphone with name "Headset (B01)" found, the device index is 3
Microphone with name "Microphone (Realtek USB2.0 Audi" found, the device index is 4
Microphone with name "Microphone (NVIDIA Broadcast)" found, the device index is 5

I would put python transcribe_audio.py --set_microphone "Microphone (Realtek USB2.0 Audi" to set the device to listen to. -or- I would put python transcribe_audio.py --set_microphone 4 to set the device to listen to.

Troubleshooting

If you encounter any issues with the tool, here are some common problems and their solutions:

• Python is not recognized as an internal or external command, operable program or batch file.
  • Make sure you have Python installed and added to your PATH.
  • If you recently installed Python, try restarting your computer to refresh the PATH environment variable.
  • Check that you installed the correct version of Python required by the application. Some applications may require a specific version of Python.
  • If you are still having issues, try running the command prompt as an administrator and running the installation again. However, only do this as a last resort and with caution, as running scripts as an administrator can potentially cause issues with the system.
• I get an error saying "No module named 'transformers'".
  • Re-run the setup.bat file.
    • If issues persist, make sure you have Python installed and added to your PATH.
    • Make sure you have the transformers module installed by running pip install transformers.
    • If you have multiple versions of Python installed, make sure you are installing the module for the correct version by specifying the Python version when running the command, e.g. python -m pip install transformers.
    • If you are still having issues, create a new issue on the repository and the developer may be able to help you fix the issue.
• Git is not recognized as an internal or external command, operable program or batch file.
  • Make sure you have Git installed and added to your PATH.
  • If you recently installed Git, try restarting your computer to refresh the PATH environment variable.
  • If you are still having issues, try running the command prompt as an administrator and running the installation again. However, only do this as a last resort and with caution, as running scripts as an administrator can potentially cause issues with the system.
• CUDA is not recognized or available.
  • Make sure you have CUDA installed. You can get it from here.
  • CUDA is only for NVIDIA GPUs. If you have an AMD GPU, you have to use the CPU model. ROCm is not supported at this time.
• [WinError 2] The system cannot find the file specified Try this fix: https://github.com/cyberofficial/Real-Time-Translation/issues/2#issuecomment-1491098222
• Translator can't pickup stream sound
  • Check out this discussion thread for a possible fix: #12 Discussion
• Error: Audio source must be entered before adjusting.
  • You need to make sure you have a microphone set up. See issue #63 for additional information.
• Error: "could not find a version that satisfies the requirement torch" (See Issue #82) )
  • Please make sure you have python 64bit installed. If you have 32bit installed, you will need to uninstall it and install 64bit. You can grab it here for windows. Windows Direct: https://www.python.org/ftp/python/3.10.9/python-3.10.9-amd64.exe Main: https://www.python.org/downloads/release/python-3109/
• Error generating captions: Please make sure the file name is in english letters. If you still get an error, please make a bug report.

Additional Information

• Models used are from OpenAI Whisper - Whisper
  • Models were fine tuned using this Documentation

Video Demonstration

Command line arguments used. --ram 6gb --record_timeout 2 --language ja --energy_threshold 500

Command line arguments used. --ram 12gb --record_timeout 5 --language id --energy_threshold 500