v2.0.0

Installation

1. Download the .exe file available below (LOKAL_2.0.0_Setup_Windows.exe).
2. Double-click it to install.
3. Use LOKAL.

It's THAT easy!

LOKAL works out-of-the-box with the default audio format (.wav) and AI models (Systran's Faster Whisper).

Usage

Detailed usage guidance, here.

System requirements

• The installer (.exe) file below is for Windows.
• You can aso use LOKAL in other systems as a pure Python app (instructions here, at the end).
• LOKAL has been tested in relatively basic computers, but speed obviously varies and the most demanding features ("diarisation" and "large" models) tend freeze in very slow computers.

OPTIONAL compatibility upgrade

LOKAL works out-of-the-box with the default audio format (.wav) and AI models (Systran's Faster Whisper), but you can also use other audio formats and AI models by running LOKAL on a computer with FFmpeg.

To install FFmpeg:

• If you have a software called "7-zip", the guidance here can help you.
• If you do not have 7-zip, the guidance here can also help you, but instead of following step 1, download "ffmpeg-master-latest-win64-gpl.zip " from here and unzip using Window's standard zip.

Alpha version

The release is an alpha version intended for testing.

Get in touch for guided pilots, or let me know of any bug if you try it independently.

v1.1.0

Installation (no coding required)

The easiest way to set LOKAL up is to:

1. Download the file available below as part of this release (LOKAL_1.0.2_Setup_Windows.exe).
2. Double-click it and install by following the instructions.
3. Use LOKAL.

It's THAT easy!

Usage guidance is available here (scroll down to the guidance section).

Prerequisites

Some features require FFmpeg. Installation guidance for FFmpeg is available here.

LOKAL will work without FFmpeg, but with reduced capacities: you will need to use .wav audios only and stick to 'Faster Whisper' models. Other audio types and models will lead to errors.

Alpha version

The release is an alpha version intended for testing.

Get in touch for guided pilots, or let us know of any bug if you try it independently.

.

.

.

Changelog (what's changed since the last version)

• New functionality added, including a GPU mode, optional timestamps, and the ability to load a prompt file for some models.
• Code was refactored in preparation for additional functionality being made available in future versions.
• README was clarified variously.

v1.0.2

Installation (no coding required)

The easiest way to set LOKAL up is to:

1. Download the installation file available as part of this release (LOKAL_1.0.2_Setup_Windows.exe).
2. Double-click it.
3. Install by following the instructions.
4. Use LOKAL.

Usage instructions available here. Please note that as per these instructions, some (but not all) features require FFmpeg.

Alpha version

The release is an alpha version intended for testing. Get in touch for guided pilots, or let us know of any bug if you try it independently.

.

.

.

Changes included in this version update

Activation

• While core functionality works regardless of whether LOKAL is activated or not, some visual features now require activation.

More visual styles available

• Users can now choose between several visual styles.

v1.0.1

Installation (no coding required)

The easiest way to set LOKAL up is:

1. Download the installation file available as part of this release (LOKAL_1.0.1_Setup_Windows.exe).
2. Double-click it.
3. Install by following the instructions.
4. Use LOKAL.

Note. Some (but not all) features require FFmpeg.

• If you do not have FFmpeg and wish to install it, guidance is available here.
• You can still use LOKAL without FFmpeg by sticking to .wav audios and Systran/faster-whisper models.

Alpha version

While it tends to work, the release is an alpha version intended for testing. Get in touch for guided pilots, or let us know of any bug if you try it independently.

.

Cumulative changes

Read it if you want a fine-graded insight into how LOKAL evolved between releases.

Multi-format support

• v1.0.1 supports numerous audio formats (all supported by FFmpeg – many, many, many formats)
• FFmpeg is required to handle audios in a format other than .wav
• That said, it is still possible to use LOKAL without FFmpeg by using .wav audios and faster-whisper models.

Model download updates

• Progress bars to keep track of model downloads render better.

Improved uninstall experience

• Uninstall script looks for and deletes models downloaded after installation.

Other

• Pop-up boxes added to warn users of key errors
• Several improvements to general logging
• Clearer README.

v1.0.0-alpha

This is the initial release of LOKAL transcriptions.

One-click install.
Users that do NOT have Python can download the "LOKAL_1.0.0_Setup_Windows.exe" file attached to this release.
Download, install, run. It is THAT easy.

Python distro.
Users that have Python on their computer might find it better to:

• Download the source code (or better yet, clone the root repository).
• Set an environment around it.
• Install dependencies listed on the file requirements.txt.
• Run LOKAL as a Python app.

Alpha version
While it tends to work, the release is an alpha version intended for testing.
Get in touch for guided pilots, or let us know of any bug if you try it independently.