video-vec2wav2-tokenizer-3 k9cli
video-vec2wav2-tokenizer-3 Version 3 - continuation shard of the video-to-AI-dataset tokenizer project. Version 2 - continuation shard of the video-to-AI-dataset tokenizer project. Production-ready pipeline (Python package video_vec2wav2_tokenizer, CLI command video2dataset) that turns a folder of videos into clean AI training datasets for speech recognition (ASR) and text-to-speech (TTS). videos ──► audio (16 kHz mono PCM) ──► whisper transcript ──► clips ──►… See the full description on the dataset page: https://huggingface.co/datasets/k9cli/video-vec2wav2-tokenizer-3.
- 类型
- dataset
- 下载量
- 305,193
- 点赞
- 1
- 访问
- public
- 文件
- 0