openvinotoolkit/openvino

[Good First Issue]: Support aten::rrelu and aten::rrelu_ for pytorch models

Open

#23,746 opened on Mar 28, 2024

View on GitHub
 (28 comments) (0 reactions) (1 assignee)C++ (3,229 forks)auto 404
category: PyTorch FEgood first issueno_stale

Repository metrics

Stars
 (10,286 stars)
PR merge metrics
 (Avg merge 14d 3h) (305 merged PRs in 30d)

Description

Context

OpenVINO component responsible for support of PyTorch models is called as PyTorch Frontend (PT FE). PT FE converts a model represented as TorchScript model to a model in OpenVINO opset.

This is a randomized operation, but it is randomized only when training, for OpenVINO we assume that models are in inference only mode, so randomization is not needed. Due to that this op is actually similar to torch.nn.LeakyReLU

What needs to be done?

  • Implement conversion rule and/or transformation to support the new operation.
  • Implement operation tests in tests/layer_tests/pytorch_tests. Please consider different data types, but keep reasonable number of test cases

Example Pull Requests

#18998

Resources

Contact points

@openvinotoolkit/openvino-pytorch-frontend-maintainers

Ticket

CVS-136484

Contributor guide

Research direction
Study the existing conversion for similar operations (e.g., aten::leaky relu) and follow the pattern. Refer to the provided examples and contribution guide.
Tech stack
pytorchpython
Domain
machine learningaibackend
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
GitC++PyTorch
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
70