huggingface / huggingface/optimum-amd

[RyzenAI] Different logs on OPT model vs CNN models

Open
#119 0 comments 0 reactions 0 assignees View on GitHub
Ryzen AI
Dominant language
Jupyter Notebook
Stars
101
Forks
21
Avg merge
58m
Merged PRs (30d)
1

Description

Ryzen SDK: Both 1.1 and 1.0.1

CNN logs:
```
[Vitis AI EP] No. of Operators : CPU 2 IPU 334 99.40%
[Vitis AI EP] No. of Subgraphs : CPU 1 IPU 1 Actually running on IPU 1
```
OPT Model logs:
```
[Vitis AI EP] No. of Operators : CPU 1404 MATMULINTEGER 73
[Vitis AI EP] No. of Subgraphs :MATMULINTEGER 73
```

* Having `Actually running on IPU` will be nice to have on OPT
* The IPU is changed to MATMULINTEGER
* Similarly in `vitisai_ep_report.json` file, for CNN, DPU ops count is written and in OPT case MATMULINTEGER count is written.

Expected behavior:
* Same logs in both cases

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.