👍🎉 First off, thanks for taking the time to contribute! 🎉👍
Our goal is to train diverse, noise-robust, multilingual models for pronunciation feedback that can run in realtime locally on consumer CPUs. We welcome all ideas from any backgrounds and are particularly open to partnerships with linguists, speech therapists, and language learners.
Checkout our general contributing guidelines here.
- Read the README.md file to understand the purpose of this repository.
- Read the DEVELOPMENT.md file to understand how to set up your local development environment. Important design decisions will also be documented here.
- Read the Code of Conduct to understand expectations while contributing to this project and take a look at the FAQ if you have any questions.
- Read through the issues to understand the current development priorities.
- We want to foster open discussion and collaboration. Feel free to open an issue to discuss early ideas or submit a pull request if you have a more concrete proposal. Training code should run locally for smaller data samples in order to be merged in. We will use our GPUs to run it on full datasets and release the results.
If you have any questions, feel free to reach out to alex@koellabs.com or aruna@koellabs.com or open an issue. Security vulnerabilities should be reported to info@koellabs.com.
By making contributions to the Koel Labs project, you agree to retain all rights, title and interest in and to your contributions and confirm that Koel Labs can use, modify, copy, and redistribute said contributions, under its choice of terms.