[BUG]--cross_entropy_fusion_impl te results in grad_norm 0
Open
bug
module: transformer engine
- Dominant language
- Python
- Stars
- 17.9k
- Forks
- 4.5k
- Avg merge
- 4d 6h
- Merged PRs (30d)
- 271
Description
After passing through this function:
https://github.com/NVIDIA/TransformerEngine/blob/6ba98d439190901fec85c2ae3c2cea235d9f2196/transformer_engine/pytorch/triton/cross_entropy.py#L359-L365
_input becomes all zeros.
Contributor guide
Assessment
This issue has not been assessed yet.