Completed
Intro
Class Central Classrooms beta
YouTube playlists curated by Class Central.
Classroom Contents
OpenAI Whisper - Robust Speech Recognition via Large-Scale Weak Supervision
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Paper overview
- 3 Collecting a large scale weakly supervised dataset
- 4 Evaluation metric issues WER
- 5 Effective robustness
- 6 Scaling laws in progress
- 7 Decoding is hacky
- 8 Code walk-through
- 9 Model architecture diagram vs code
- 10 Transcription task
- 11 Loading the audio, mel spectrograms
- 12 Language detection
- 13 Transcription task continued
- 14 Suppressing token logits
- 15 Voice activity detection
- 16 Decoding and heuristics
- 17 Outro
