Welcome!

[wasertech]

This project aims to democratize better scorer for Acoustic Speech Recognition (ASR).

For now it's in version alpha 1; meaning we focus on functionality at the expense of everything else. Including proper documentation.

Although it's temporarily inconvenient to use, it is a crucial phase of the project were we make sure the core project is stable and able to fully complete a pipeline run for ASR.

Meaning TranScorerLM achieves:

• Training
• Evaluation
• Test Metrics
• Model Inference

Once we get a few annoying bugs out of this release, we'll focus on introducing proper documentation but also some basic automatic tests and more.

Feel free to use this space to share and improve your own models.

Guidelines

When chatting, please keep the following guidelines in mind:

• Be respectful to others and their ideas
• Avoid making a new topic when once seems appropriate