NVIDIA / NVIDIA/TensorRT

`polygraphy run` failed on Polygraphy Version: 0.49.20. when will update `FormattedArray` to support reformated model input/output tensor, such as fp16:hwc8

Open
#4,685 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Module:Polygraphy
Dominant language
C++
Stars
13.4k
Forks
2.4k
Avg merge
5d 3h
Merged PRs (30d)
2

Description

polygraphy: print or validate_output for struct FormattedArray is unsupported, while my trt engine's output is reformated to fp16:hwc8.

error log from polygraphy run of my trt:
Function: dtype is unsupported for objects of type: FormattedArray
Function: mean is unsupported for objects of type: FormattedArray
Function: size is unsupported for objects of type: FormattedArray

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 polygraphy run with a TensorRT engine whose output is formatted as fp16:hwc8, then trace how FormattedArray is handled by the print and validate_output paths. Done means these paths no longer fail on the reported dtype, mean, and size operations.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Active
Clarity
Needs clarification
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.