{joao_leonardi.melo, enzo.goncalves, joao_vinicius.carvalho}@somosicev.com
Este projeto está sendo desenvolvido inicialmente para a plataforma x86-64 amd64 Linux, ainda não tem previsão para o port de outras plataformas.
Hardware principal usado no desenvolvimento atual:
OS: Linux x86-64 amd4 Kernel 6.2 Genric
CPU: 12th Gen Intel® Core™ i5-12450H × 12
GPU: NVIDIA RTX 3050 Mobile Cuda cores: 2048 Neural Optimization
Mem: 24.0 GiB
TGP: 95W- Linux
- Windows (WSL)
- MacOS
- Aceleração via CPU (intel)
- Aceleração via GPU (nvidia)
- AMD graphics OpenCL.
volts ⚡️
|
+--- engines/
| |
| +--- weights/
|
+--- performance/inference/
| |
| +--- utils/Utils.py
| +--- board.svg
| +--- main.py
|
+--- pyproject.toml
+--- README.md
+--- requeriments.txt
+--- ubuntu_deps.txtPapers:
Chess AI: Competing Paradigms for Machine Intelligence:
https://arxiv.org/pdf/2109.11602.pdf
Neural Networks for Chess
https://arxiv.org/pdf/2209.01506.pdf
Bibliotecas
Autor: João Leonardi
Neural Network Library for C lang 🧠
A collection of library for Deep Learning
https://github.com/rexionmars/neural-network-library
CTorch 🔥
A next library stb-style header-only library for Neural Networks
https://github.com/rexionmars/ctorch
João Leonardi da Silva Melo
João Vinícius Castelo branco
Enzo Morais Gonçalves
Email:
joao_leonardi.melo@somosicev.com
joao_vinicius.carvalho@somosicev.com
enzo.goncalves@somosicev.com