🚀 SymbolicRegressionGPU.jl

💻 PSRN (Parallel Symbolic Regression Network) enhanced SymbolicRegression.jl via faster, large-scale parallel symbolic evaluations on GPUs. Based on SymbolicRegression.jl.

Quickstart

📥 1. clone this repo

git clone https://github.com/x66ccff/SymbolicRegressionGPU.jl

📦 2. Install

julia ]
(@v1.1x) pkg> activate .
(SymbolicRegressionGPU) pkg> instantiate
(SymbolicRegressionGPU) pkg> resolve

# (deprecated)
# conda activate ./.CondaPkg/.pixi/envs/default
# (default) $ uv pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu126

using CondaPkg
CondaPkg.add("pip")
/home/kent/.julia/environments/v1.11/.CondaPkg/.pixi/envs/default/bin/pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu126

🏃‍♂️ 3. Run

# Note: only supports one thread now
julia example.jl -t 16,2

📚 Citing

To cite this fork SymbolicRegressionGPU.jl, please use the following BibTeX entry:

@misc{SymbolicRegressionGPU.jl,
  author = {
    Ruan, Kai AND
    Cranmer, Miles AND
    Sun, Hao
  },
  title = {SymbolicRegressionGPU.jl: PSRN enhanced SymbolicRegression.jl via fast, large-scale parallel symbolic evaluations on GPUs}, 
  year = {2024},
  url = {https://github.com/x66ccff/SymbolicRegressionGPU.jl}
}

@misc{cranmerInterpretableMachineLearning2023,
    title = {Interpretable {Machine} {Learning} for {Science} with {PySR} and {SymbolicRegression}.jl},
    url = {http://arxiv.org/abs/2305.01582},
    doi = {10.48550/arXiv.2305.01582},
    urldate = {2023-07-17},
    publisher = {arXiv},
    author = {Cranmer, Miles},
    month = may,
    year = {2023},
    note = {arXiv:2305.01582 [astro-ph, physics:physics]},
    keywords = {Astrophysics - Instrumentation and Methods for Astrophysics, Computer Science - Machine Learning, Computer Science - Neural and Evolutionary Computing, Computer Science - Symbolic Computation, Physics - Data Analysis, Statistics and Probability},
}

🎉 Enjoy your symbolic regression journey with SymbolicRegressionGPU.jl! 🎉