RuntimeError: Dataset scripts are no longer supported, but found spgispeech.py
bugneeds triage
## 🐛 Bug
The example code provided in documentation is not working
### To Reproduce
Steps to reproduce the behaviour:
1. uv venv
2. uv source .venv/bin/activate
3. python
4. from datasets import load_dataset; spgi = load_dataset("kensho/spgispeech", "S", use_auth_token=True)
5. See error: ... RuntimeError: Dataset scripts are no longer supported, but found spgispeech.py
#### Code sample
from datasets import load_dataset
spgi = load_dataset("kensho/spgispeech", "S", use_auth_token=True)
\# see structure
print(spgi)
\# load audio sample on the fly
audio_input = spgi["train"][0]["audio"] # first decoded audio sample
transcription = spgi["train"][0]["text"] # first transcription
### Expected behavior
Output first decoded audio sample and it's transcription
### Environment
- uv version: uv 0.8.3 (Homebrew 2025-07-24)
- OS: macOS Sequila 15.6
- Machine: Macbook M2
关闭于 2025-08-06 1 条评论