Open-source PyTorch toolkit for conversational AI, covering speech recognition, text-to-speech, speaker recognition and audio processing.
SpeechBrain is released under the Apache 2.0 license and has more than 200 training recipes for speech, audio and language tasks. These include speech recognition, enhancement and separation, text-to-speech, speaker recognition, speech-to-speech translation and spoken language understanding. It also covers audio work such as vocoding, sound event detection and beamforming, and language models from n-gram to LLM. Each recipe puts its hyperparameters in a single YAML file and trains from one Python script, so researchers can swap in custom models, losses and training loops. Pretrained models are on Hugging Face, and the toolkit installs with pip install speechbrain. SpeechBrain 1.0 was released in 2024.
Speech-to-Text
FOUNDED
2021
Is this your company? To claim this listing or update its details, contact us at sunil@uniocommunity.com
