ESPnet Notebooks

Notebooks for ESPnet: short demos of what the toolkit does, and the material from the CMU speech courses.

Each badge is that notebook, executed cell by cell every Sunday against the release it pins. Nothing joins a table until it runs there — what green means, and what it does not.

Demos

One per task, flat in Demos/, each short enough to read in a sitting.

Notebook		What it does
asr_demo.ipynb		Transcribe speech with OWSM-CTC, and let it work out the language
asr_streaming_demo.ipynb		Watch the words appear while the audio is still arriving
st_demo.ipynb		Translate English speech into German, French and Chinese — the same model
s2t_align_demo.ipynb		Line text up with the audio it was said in, and score how well they agree
tts_demo.ipynb		Type a sentence, hear it spoken — one English voice, then 128 of them
enh_demo.ipynb		Pull speech out of noise, and measure how much it helped
spk_demo.ipynb		Turn a voice into a vector, and score two recordings against each other
codec_demo.ipynb		Compress a waveform to a few integers a frame, rebuild it, count the bits
sds_demo.ipynb		Speech in, speech out, with a language model thinking in between

Demos/unmaintained/ holds what was here before — the oldest from 2021, most of it no longer running, kept because some of it is still the only written record of how a thing was done. It carries no badge and nothing checks it.

Demos/README.md has the naming rule and what a demo owes the reader.

Courses

CMU Speech Technology 26S

CMU 11492/11692/18495, Speech Technology for Conversational AI, Spring 2026 — the demonstration notebooks, with the graded exercises removed, in Courses/CMUSpeechTechnology26S/.

Notebook		What it does
speaker_verification.ipynb		Speaker embeddings with ESPnet-SPK, verification, and a simple diarization
speech_enhancement.ipynb		Enhancement and separation, scored with VERSA and a pretrained ASR model
text_to_speech.ipynb		Single-speaker and multi-speaker synthesis, and VERSA scores
neural_codec.ipynb		Three pretrained neural codecs and the bitrate trade between them
speech_translation.ipynb		Offline and simultaneous speech translation with ESPnet-ST-v2
speech_recognition.ipynb		Fine-tune OWSM on one language of FLEURS with the ESPnet3 trainer

Earlier courses

Courses/ also holds the material from CMU 11492/11692 Spring 2023, and 11751/18781 Fall 2021 and Fall 2022. Nothing runs them and nothing has checked them for years; they are kept as a record of how these things were taught and done. Open one expecting to fix it before it works.