qts8n/darknet

Darknet fork with QoL improvements and custom Docker images

Darknet

Object detection neural network framework

Darknet is an open source neural network framework written in C and CUDA. It is fast, easy to install, and supports CPU and GPU computation.

This fork adds the possibility to pass a callback function in the training pipeline for logging purposes and fixes a bunch of memory leaks in the training process.

It adds the support of recent RTX architectures and Dockerfiles to use as a base of your projects.

Dockerfiles are based on on devel tags of qts8n/cuda-python image. It has cuda-9.1 and cuda-10.2 tags with support of CUDA 9.1 and CUDA 10.2 versions respectively.

References

see: Darknet github

see: Darknet project website

see: qts8n/cuda-python docker-hub

see: nvidia/cuda docker-hub