huggingface / huggingface/datasets
train_dataset does not implement __len__
Open
- Dominant language
- Python
- Stars
- 22k
- Forks
- 3.4k
- Avg merge
- 5d 7h
- Merged PRs (30d)
- 17
Description
when train using data precessored by the datasets, I get follow warning and it leads to that I can not set epoch numbers:
`ValueError: The train_dataset does not implement __len__, max_steps has to be specified. The number of steps needs to be known in advance for the learning rate scheduler.`
Contributor guide
Assessment
This issue has not been assessed yet.