mirror of
https://github.com/zebrajr/pytorch.git
synced 2025-12-06 12:20:52 +01:00
Digging further the problem at https://github.com/UKPLab/sentence-transformers/pull/3261, it boils down to this expensive loop over a torch tensor. Looping over a list, like in RandomSampler, solves the issue. Pull Request resolved: https://github.com/pytorch/pytorch/pull/149126 Approved by: https://github.com/divyanshk, https://github.com/cyyever |
||
|---|---|---|
| .. | ||
| _utils | ||
| datapipes | ||
| __init__.py | ||
| backward_compatibility.py | ||
| dataframes_pipes.ipynb | ||
| dataloader.py | ||
| dataset.py | ||
| distributed.py | ||
| graph_settings.py | ||
| graph.py | ||
| sampler.py | ||
| standard_pipes.ipynb | ||
| typing.ipynb | ||