https://github.com/sowson/darknet
Darknet on OpenCL Convolutional Neural Networks on OpenCL on Intel & NVidia & AMD & Mali GPUs for macOS & GNU/Linux & Windows & FreeBSD
https://github.com/sowson/darknet
amd c cpu darknet freebsd gpu intel linux macos macosx nvidia opencl windows
Last synced: 6 months ago
JSON representation
Darknet on OpenCL Convolutional Neural Networks on OpenCL on Intel & NVidia & AMD & Mali GPUs for macOS & GNU/Linux & Windows & FreeBSD
- Host: GitHub
- URL: https://github.com/sowson/darknet
- Owner: sowson
- License: other
- Created: 2018-07-26T03:45:14.000Z (about 7 years ago)
- Default Branch: master
- Last Pushed: 2025-04-02T23:35:54.000Z (6 months ago)
- Last Synced: 2025-04-12T23:43:02.405Z (6 months ago)
- Topics: amd, c, cpu, darknet, freebsd, gpu, intel, linux, macos, macosx, nvidia, opencl, windows
- Language: C
- Homepage: http://pjreddie.com/darknet/
- Size: 32.1 MB
- Stars: 192
- Watchers: 17
- Forks: 33
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Project Status: Work in progress!
- Many ideas I used and focused on https://iChess.io Chess and Chess960 Gameplay! Not yet fully Integrated! :D.
# iChess.io v7.27 Engine and AI/ML
- UCI engine to use build as fallows:
mkdir build && cd build && cmake -DDARKNET_ENG_CHESS=1 .. && make && cp darknet /your/engine/path
- in the engine path you have to put chess.cfg and chess.weights in the Linux or macOS may use RAMDisk.
- Linux: sudo mount -t tmpfs -o rw,size=4096M tmpfs /your/disk.ram/
- macOS: diskutil erasevolume HFS+ "/your/disk.ram" `hdiutil attach -nomount ram://8388608`
Using RAM disk in general is good to safe your HDD or SDD with often transactions :D.
# Dependencies:
- json (https://github.com/nlohmann/json)
- libchess (https://github.com/sowson/libchess)
# Build on macOS or Ubuntu 20.04
Step by step in command prompt guide: https://iblog.isowa.io/2018/05/26/darknet-training
# Build on Windows 10 or 11 x64
Step by step experimental guide: https://iblog.isowa.io/2021/11/20/darknet-on-opencl-on-windows-11-x64
# Take a look 4 x GPUs on macOS (click to see video)
[](https://www.youtube.com/watch?v=W6VOLjgwKNI)
# CLBlast instead of clBLAS for GEMM
git apply patches/clblast.patch
# Darknet-vNext (Improved CUDA DarkNet)
[DarkNet-vNext Link](https://github.com/sowson/darknet-vNext) If you are looking for engine that has all the same functions, but it is FASTER!
# OpenCV 4
This engine runs on OpenCV v4! But, OpenCV v3 is also fine!
# YOLO4 on OpenCL
YOLO4 elements are supported, remember in CFG file to use [yolo4] instead of [yolo] to make it work!
# YOLO1, YOLO2, YOLO3 on OpenCL
[](https://www.youtube.com/watch?v=o-PV3vmfP-0)
https://iblog.isowa.io/2020/07/02/the-multi-gpu-set-idea
[](https://www.youtube.com/watch?v=_dNYNYHXHHo)
https://iblog.isowa.io/2020/06/22/gpu-opencl-fine-tuning-problem-solution
https://iblog.isowa.io/2020/05/31/ph-d-hanna-hackintosh-is-ready
[](https://www.youtube.com/watch?v=qfCWYVnJrjQ)
https://iblog.isowa.io/2020/04/29/darknet-in-opencl-on-beagleboard-ai
[](https://www.youtube.com/watch?v=exuPfFtbwgU)
https://iblog.isowa.io/2020/03/03/is-opencl-beats-cuda
https://iblog.isowa.io/2020/03/02/hania-pc-well-it-needs-macos
https://iblog.isowa.io/2020/02/08/pc-for-phd-studies
https://iblog.isowa.io/2020/01/04/gpu-opencl-fine-tuning-problem
https://iblog.isowa.io/2019/12/29/darknet-cuda-vs-opencl-and-cpu-vs-nvidia-vs-amd
https://iblog.isowa.io/2019/11/05/gpu-computing-on-opencl
https://iblog.isowa.io/2019/08/18/the-fastest-darknet-in-opencl-on-the-planet
https://iblog.isowa.io/2019/02/02/darknet-in-opencl-on-asus-thinker-board-s
[](https://www.youtube.com/watch?v=Mxw7XkFBFPc)
https://iblog.isowa.io/2018/08/01/darknet-in-opencl
https://iblog.isowa.io/2018/05/26/darknet-training
Thanks!