NVIDIA / NVIDIA/TensorRT-RTX

[Feature Request] Support NonZero op which is supported by original tensorrt

Open
#36 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement
Dominant language
No language data
Stars
109
Forks
19
PR merge metrics
No merged PRs in 30d

Description

[07/08/2026-12:08:04] [TRT] [E] [tacticSupply.cpp::getSupportedFormats::48] Error Code 2: Internal Error (Assertion !layerImpls.empty() failed. /model/model/transformer_module/decoder/NonZero has no runners that implement it In getSupportedFormats at /_src/optimizer/builder/tacticSupply.cpp:48)

I'm using 1.5.0.114
with cuda 13.2

This was a non issue in Tensor RT 11

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

Start by reproducing the reported failure in the transformer decoder at the NonZero node with TensorRT-RTX 1.5.0.114 and CUDA 13.2, then compare its behavior with TensorRT 11. Done means the reported model no longer fails with the getSupportedFormats assertion and NonZero produces the expected result.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
45/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.