yolov8_tf-serving

yolov8_tf-serving is a project designed to convert YOLOv8 models into a format compatible with TensorFlow Serving, enabling seamless deployment of these models in production environments.

Getting Started

• Clone the repository

git clone https://github.com/Kawaeee/yolov8_tf-serving.git
cd yolov8_tf-serving/

• Build Docker image

docker build -t yolov8conv .

• Access Docker container bash shell

# CPU
docker run -it -v $(pwd):/data --rm yolov8conv /bin/bash

# GPU
docker run -it -v $(pwd):/data --gpus all --rm yolov8conv /bin/bash

• Run run.sh with .pt model file in mounted directory

bash /app/run.sh <yolov8-model.pt>

# Example:
wget -O /data/yolov8l.pt https://github.com/ultralytics/assets/releases/download/v0.0.0/yolov8l.pt
bash /app/run.sh /data/yolov8l.pt

• After obtaining the converted model, transfer the contents from the "output" directory to the "demo/models". Then, simply follow the instructions outlined in the next steps.