https://github.com/kyflores/kv260-vitis-flow

Self contained example of training, inspecting, and quantizing a model for the KV260 DPU
https://github.com/kyflores/kv260-vitis-flow

Last synced: 29 days ago
JSON representation

Self contained example of training, inspecting, and quantizing a model for the KV260 DPU

• Host: GitHub
• URL: https://github.com/kyflores/kv260-vitis-flow
• Owner: kyflores
• License: mit
• Created: 2023-10-02T03:19:13.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-12-19T01:29:58.000Z (about 1 year ago)
• Last Synced: 2024-11-09T15:05:51.290Z (3 months ago)
• Language: C
• Size: 84 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# kv260-vitis-flow

Self contained example of training, inspecting, and quantizing a model for the KV260 DPU

https://github.com/Xilinx/Vitis-AI/tree/v3.5/src/vai_library

## Compiling the model

After quantizing and exporting the model still needs to be compiled for the specific DPU arch

Do this after you get your xmodel from pt2

https://xilinx.github.io/kria-apps-docs/kv260/2022.1/build/html/docs/smartcamera/docs/customize_ai_models.html

arch.json (get it from xdputil query)

```

{

    "fingerprint":"0x101000016010406"

}

```

Then compile to produce deploy.xmodel and md5sum.txt

```

vai_c_xir -x /path/to/xmodel -a /path/to/arch.json

```

idk about compatibility guarantees but model quantized and compiled by Vitis 3.5

worked on the KV260 image that's on Vitis runtime/library 2.5