kospeech and openspeech
These are ecosystem siblings—OpenSpeech is a language-agnostic PyTorch-Lightning framework that generalized and superseded KoSpeech's Korean-specific PyTorch approach, with both sharing Hydra for configuration management.
About kospeech
sooftware/kospeech
Open-Source Toolkit for End-to-End Korean Automatic Speech Recognition leveraging PyTorch and Hydra.
Supports multiple end-to-end architectures (Deep Speech 2, LAS, Transformer, Jasper, Conformer, RNN-Transducer) with configurable components like attention mechanisms and joint CTC-Attention training. Built around Hydra for hierarchical configuration management and optimized for the KsponSpeech corpus with established preprocessing pipelines and baseline comparisons.
About openspeech
openspeech-team/openspeech
Open-Source Toolkit for End-to-End Speech Recognition leveraging PyTorch-Lightning and Hydra.
Provides reference implementations of 20+ ASR architectures (DeepSpeech2, RNN-Transducer, ContextNet, etc.) with support for encoder-decoder and transducer-based decoding strategies including beam search. Includes multi-language recipes for English, Mandarin, and Korean with pre-configured datasets (LibriSpeech, AISHELL-1, KsponSpeech) and modular audio processing with SpecAugment, mel-spectrograms, and MFCC feature extraction. Leverages Hydra's hierarchical configuration system to enable hardware-agnostic training across multi-GPU, TPU, and mixed-precision setups without code changes.
Scores updated daily from GitHub, PyPI, and npm data. How scores work