Mixed precision training slow
Open
Nobody has claimed this yet.
BERT
- Dominant language
- Python
- Stars
- 9k
- Forks
- 1.5k
- Avg merge
- 2d 4h
- Merged PRs (30d)
- 3
Description
Hi,
I'm trying to fine-tuning bert using Bert fine-tuning.
My problem is: after using apex, the GPU memory usage is reduced, but the training time is about 1.3 times before.
My GPU is V100(16G, CUDA9, CUDNN7), Pytorch version is 1.0.
Is it a problem with my hardware?
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reviewing the linked Kaggle BERT fine-tuning notebook and the reported Apex, PyTorch, CUDA, cuDNN, and V100 versions. Compare training with and without mixed precision and identify whether the slowdown is reproducible; done means establishing a clear cause or documenting that it is hardware or configuration dependent.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100