Other Speech-to-Text frameworks

[egorsmkv]

We can use other e2e frameworks to train STT models:

• wenet: https://github.com/wenet-e2e/wenet
• wav2vec2-conformer
• wavlm
• Scribosermo: https://gitlab.com/Jaco-Assistant/Scribosermo
• icefall
• SpeechT5
• CMU Sphinx
• flashlight
• contextnet (nvidia)

[egorsmkv]

• emformer
• espnet

[egorsmkv]

• https://github.com/thu-spmi/CAT

[egorsmkv]

data2vec 2.0 appeared