UserWarning: Grad strides do not match bucket view strides.
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 198
- Forks
- 21
- PR merge metrics
- No merged PRs in 30d
Description
Have you encountered such problems during training?：
.conda/envs/seggpt/lib/python3.8/site-packages/torch/autograd/init.py:173: UserWarning: Grad strides do not match bucket view strides. This may indicate grad was not created according to the gradient layout contract, or that the param's strides changed since DDP was constructed. This is not an error, but may impair performance.
grad.sizes() = [1, 1, 1024], strides() = [2409472, 1024, 1]
bucket_view.sizes() = [1, 1, 1024], strides() = [1024, 1024, 1] (Triggered internally at ../torch/csrc/distributed/c10d/reducer.cpp:312.)
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
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 reproducing the training run that emits the warning and inspect the gradient and bucket strides reported by torch.autograd and the DDP reducer at ../torch/csrc/distributed/c10d/reducer.cpp:312. Done means determining whether the warning reflects a repository issue or an expected performance warning, and recording the cause and resolution.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100