{"id":17632759,"url":"https://github.com/brosnanyuen/raybnn_optimizer","last_synced_at":"2025-10-07T16:05:10.400Z","repository":{"id":201117926,"uuid":"707004588","full_name":"BrosnanYuen/RayBNN_Optimizer","owner":"BrosnanYuen","description":"Gradient Descent Optimizers and Genetic Algorithms using GPUs, CPUs, and FPGAs via CUDA, OpenCL, and oneAPI ","archived":false,"fork":false,"pushed_at":"2024-06-05T18:07:30.000Z","size":104,"stargazers_count":3,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-09-20T14:39:06.372Z","etag":null,"topics":["arrayfire","cuda","genetic-algorithm","genetic-algorithms","gpu","gpu-computing","gradient","gradient-descent","parallel","parallel-computing","raybnn","rust"],"latest_commit_sha":null,"homepage":"","language":"Rust","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/BrosnanYuen.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2023-10-19T03:12:05.000Z","updated_at":"2024-10-17T12:38:59.000Z","dependencies_parsed_at":"2024-10-23T07:19:03.739Z","dependency_job_id":"2aeaf764-0fbf-4345-ae8b-2eb00db78d3b","html_url":"https://github.com/BrosnanYuen/RayBNN_Optimizer","commit_stats":{"total_commits":165,"total_committers":2,"mean_commits":82.5,"dds":0.0060606060606061,"last_synced_commit":"a2702d9d0dafa643f4d4f64fd8cca76358a21790"},"previous_names":["brosnanyuen/raybnn_optimizer"],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/BrosnanYuen/RayBNN_Optimizer","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BrosnanYuen%2FRayBNN_Optimizer","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BrosnanYuen%2FRayBNN_Optimizer/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BrosnanYuen%2FRayBNN_Optimizer/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BrosnanYuen%2FRayBNN_Optimizer/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/BrosnanYuen","download_url":"https://codeload.github.com/BrosnanYuen/RayBNN_Optimizer/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BrosnanYuen%2FRayBNN_Optimizer/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":278802847,"owners_count":26048582,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-10-07T02:00:06.786Z","response_time":59,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["arrayfire","cuda","genetic-algorithm","genetic-algorithms","gpu","gpu-computing","gradient","gradient-descent","parallel","parallel-computing","raybnn","rust"],"created_at":"2024-10-23T01:45:38.243Z","updated_at":"2025-10-07T16:05:10.385Z","avatar_url":"https://github.com/BrosnanYuen.png","language":"Rust","funding_links":[],"categories":[],"sub_categories":[],"readme":"# RayBNN_Optimizer\n\nGradient Descent Optimizers and Genetic Algorithms using GPUs, CPUs, and FPGAs via CUDA, OpenCL, and oneAPI \n\n* ADAM\n* SGD\n* Genetic\n* Random Search\n\n\n\n\n# Install Arrayfire\n\nInstall the Arrayfire 3.9.0 binaries at [https://arrayfire.com/binaries/](https://arrayfire.com/binaries/)\n\nor build from source\n[https://github.com/arrayfire/arrayfire/wiki/Getting-ArrayFire](https://github.com/arrayfire/arrayfire/wiki/Getting-ArrayFire)\n\n\n\n\n# Add to Cargo.toml\n```\narrayfire = { version = \"3.8.1\", package = \"arrayfire_fork\" }\nrayon = \"1.10.0\"\nnum = \"0.4.3\"\nnum-traits = \"0.2.19\"\nhalf = { version = \"2.4.1\" , features = [\"num-traits\"] }\nRayBNN_Optimizer = \"2.0.1\"\n```\n\n# List of Examples\n\n\n# Optimizing values for a loss function\n```\n\n//Define Starting Point for optimization\nlet x0_cpu = vec![0.1, 0.4, 0.5,   -1.2, 0.7];\nlet x0_dims = arrayfire::Dim4::new(\u0026[1, x0_cpu.len() as u64, 1, 1]);\nlet x0 = arrayfire::Array::new(\u0026x0_cpu, x0_dims);\n\n//Define the loss function\nlet y_cpu = vec![-1.1, 0.4, 2.0,    2.1, 4.0];\nlet y = arrayfire::Array::new(\u0026y_cpu, x0_dims);\n\n//Define the loss function\nlet loss = |yhat: \u0026arrayfire::Array\u003cf64\u003e| -\u003e arrayfire::Array\u003cf64\u003e {\n    RayBNN_Optimizer::Continuous::Loss::MSE(yhat, \u0026y)\n};\n\n//Define the gradient of the loss function\nlet loss_grad = |yhat: \u0026arrayfire::Array\u003cf64\u003e| -\u003e arrayfire::Array\u003cf64\u003e {\n    RayBNN_Optimizer::Continuous::Loss::MSE_grad(yhat, \u0026y)\n};\n\n\nlet mut point = x0.clone();\nlet mut direction = -loss_grad(\u0026point);\nlet mut mt = arrayfire::constant::\u003cf64\u003e(0.0,direction.dims());\nlet mut vt = arrayfire::constant::\u003cf64\u003e(0.0,direction.dims());\n\nlet single_dims = arrayfire::Dim4::new(\u0026[1,1,1,1]);\nlet mut alpha = arrayfire::constant::\u003cf64\u003e(1.0,single_dims);\n\nlet alpha_max = arrayfire::constant::\u003cf64\u003e(1.0,single_dims);\n\nlet rho = arrayfire::constant::\u003cf64\u003e(0.1,single_dims);\n\n//Create alpha values to sweep through\nlet v = 30;\nlet alpha_vec = RayBNN_Optimizer::Continuous::LR::create_alpha_vec::\u003cf64\u003e(v, 1.0, 0.5);\n\n\nlet beta0 = arrayfire::constant::\u003cf64\u003e(0.9,single_dims);\nlet beta1 = arrayfire::constant::\u003cf64\u003e(0.999,single_dims);\n\n//Optimization Loop\nfor i in 0..120\n{\n    alpha = alpha_max.clone();\n    //Automatically Determine Optimal Step Size using BTLS\n    RayBNN_Optimizer::Continuous::LR::BTLS(\n        loss\n        ,loss_grad\n        ,\u0026point\n        ,\u0026direction\n        ,\u0026alpha_vec\n        ,\u0026rho\n        ,\u0026mut alpha\n    );\n\n    //Update current point\n    point = point.clone()  + alpha*direction.clone();\n    direction = -loss_grad(\u0026point);\n\n\n\n    //Use ADAM optimizer\n    RayBNN_Optimizer::Continuous::GD::adam(\n        \u0026beta0\n        ,\u0026beta1\n        ,\u0026mut direction\n        ,\u0026mut mt\n        ,\u0026mut vt\n    );\n\n}\n\n```\n\n\n# Types of Loss Functions\n```\nlet mut cross_entropy = RayBNN_Optimizer::Continuous::Loss::softmax_cross_entropy(\u0026Yhat,\u0026Y);\nlet mut cross_entropy_grad = RayBNN_Optimizer::Continuous::Loss::softmax_cross_entropy_grad(\u0026Yhat,\u0026Y);\nlet mut cross_entropy = RayBNN_Optimizer::Continuous::Loss::sigmoid_cross_entropy(\u0026Yhat,\u0026Y);\nlet mut cross_entropy_grad = RayBNN_Optimizer::Continuous::Loss::sigmoid_cross_entropy_grad(\u0026Yhat,\u0026Y);\nlet mut MAE = RayBNN_Optimizer::Continuous::Loss::MAE(\u0026Yhat,\u0026Y);\nlet mut MSE = RayBNN_Optimizer::Continuous::Loss::MSE(\u0026Yhat,\u0026Y);\nlet MSE_grad = RayBNN_Optimizer::Continuous::Loss::MSE_grad(\u0026Yhat,\u0026Y);\nlet mut RMSE = RayBNN_Optimizer::Continuous::Loss::RMSE(\u0026Yhat,\u0026Y);\n```\n\n\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbrosnanyuen%2Fraybnn_optimizer","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fbrosnanyuen%2Fraybnn_optimizer","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbrosnanyuen%2Fraybnn_optimizer/lists"}