pytorch / pytorch/executorch

[Vulkan] Make partitioning dtype aware

Open
#13,084 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

module: vulkan
Dominant language
Python
Stars
5k
Forks
1.2k
Avg merge
2d 10h
Merged PRs (30d)
581

Description

🚀 The feature, motivation and pitch

Context

Currently, Vulkan's partitioner does not account for the dtype of expected input/output tensors participating in an op. This can lead to a situation where the runtime receives dtypes for an op that it can't handle.

Alternatives

No response

Additional context

No response

RFC (Optional)

No response

cc @manuelcandales @cbilgin

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

Begin by locating the Vulkan partitioner and tracing how it determines expected input and output tensor dtypes; the issue does not name a file or test. Done means partitioning accounts for those dtypes so the runtime does not receive unsupported dtypes.

Written by the indexing model from the issue text.

Assessment

Domain
computer-graphics
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
28/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.