YOLOv5s model resources

Code and deployment steps: https://github.com/APLUX-Agora/yolov5s-demo

SoC	Precisions	Compiled QNN
QCS6490	INT8, W8A16	2.36
QCS8550	FP16, INT8, W8A16	2.48
QCS8625 / IQ8 / IQ9	FP16, W8A16	2.36

Download

export AGORA_WORKSPACE="$HOME/my-yolo-demo"
export AGORA_SOC='QCS8550'
hf download APLUX-Agora/YOLOv5s \
  --include "targets/$AGORA_SOC/*" --local-dir "$AGORA_WORKSPACE/resources"

Set the two global variables directly in your device terminal; choose any workspace path. The GitHub README provides the same five deployment steps. Each targets/<SoC>/ contains model.json and one model.bin per precision. Input: 640 x 640 RGB. Output: [1, 25200, 85], COCO 80 classes.

Binary contents are unchanged from the ModelFarm sources. See SHA256SUMS and NOTICE.md. License: AGPL-3.0; see LICENSE. Runtime SDKs are installed separately.

After inference on a cloud device, switch to the remote desktop and open result.jpg in your workspace. See the GitHub README for the run command.

Downloads last month: -; Downloads are not tracked for this model. How to track