pytorch / pytorch/tutorials

Why libtorch tensor value assignment takes so much time?

Open
#1,534 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

question Tensors
Dominant language
Python
Stars
9.3k
Forks
4.4k
Avg merge
1d 21h
Merged PRs (30d)
4

Description

I just assign 10000 values to a tensor:

clock_t start = clock();
torch::Tensor transform_tensor = torch::zeros({ 10000 });
for (size_t m = 0; m < 10000 m++)
	transform_tensor[m] = int(m);
clock_t finish = clock();

And it takes 0.317s. If I assign 10,000 to an array or a vector, the time cost will be less.
Why tensor takes so much time? Can the time cost be decreased?

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Use the embedded C++ loop and its libtorch tensor assignment as the entry point, then reproduce the timing with the same 10,000-element workload. Check the relevant tensor indexing and assignment behavior before comparing alternatives. Done means documenting the cause of the measured cost and whether a supported way to reduce it exists.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.