{"id":31227633,"url":"https://github.com/lordmathis/cudanet","last_synced_at":"2025-09-22T05:02:11.116Z","repository":{"id":221415579,"uuid":"754290422","full_name":"LordMathis/CUDANet","owner":"LordMathis","description":"Convolutional Neural Network inference library running on CUDA","archived":false,"fork":false,"pushed_at":"2024-09-10T19:48:51.000Z","size":7156,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":2,"default_branch":"main","last_synced_at":"2024-09-11T01:17:10.633Z","etag":null,"topics":["convolutional-neural-networks","cpp","cuda","pytorch"],"latest_commit_sha":null,"homepage":"","language":"Cuda","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/LordMathis.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":"2024-02-07T19:06:29.000Z","updated_at":"2024-09-10T19:48:54.000Z","dependencies_parsed_at":"2024-03-07T21:31:09.654Z","dependency_job_id":"b1709f14-1f08-40eb-9a75-8175bffb66a1","html_url":"https://github.com/LordMathis/CUDANet","commit_stats":null,"previous_names":["lordmathis/cudanet"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/LordMathis/CUDANet","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LordMathis%2FCUDANet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LordMathis%2FCUDANet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LordMathis%2FCUDANet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LordMathis%2FCUDANet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/LordMathis","download_url":"https://codeload.github.com/LordMathis/CUDANet/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LordMathis%2FCUDANet/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":276348331,"owners_count":25626605,"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":"2025-09-22T02:00:08.972Z","response_time":79,"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":["convolutional-neural-networks","cpp","cuda","pytorch"],"created_at":"2025-09-22T05:00:46.427Z","updated_at":"2025-09-22T05:02:11.102Z","avatar_url":"https://github.com/LordMathis.png","language":"Cuda","readme":"# CUDANet\n\nConvolutional Neural Network inference library running on CUDA.\n\n## Quickstart Guide\n\n**requirements**\n- [cmake](https://cmake.org/)\n- [CUDA](https://developer.nvidia.com/cuda-downloads)\n- [Google Test](https://github.com/google/googletest) (for testing only)\n\n**build**\n\n```sh\nmkdir build\ncd build\ncmake -S .. -DCMAKE_CUDA_ARCHITECTURES=75  # Replace with you cuda architecture\nmake\n```\n\n**build and run tests**\n\n```sh\nmake test_main\n./test/test_main\n```\n\n### Create Layers and Model\n\n```cpp\nCUDANet::Model *model =\n    new CUDANet::Model(inputSize, inputChannels, outputSize);\n\n// Conv2d\nCUDANet::Layers::Conv2d *conv2d = new CUDANet::Layers::Conv2d(\n    inputSize, inputChannels, kernelSize, stride, numFilters,\n    CUDANet::Layers::Padding::VALID,\n    CUDANet::Layers::ActivationType::NONE\n);\n\nif (setWeights) {\n    conv2d-\u003esetWeights(getConv1Weights().data());\n}\nmodel-\u003eaddLayer(\"conv1\", conv2d);\n```\n\n### Sequential and Functional API\n\nRun prediction by passing the input through the layers in the order they have been added.\n\n```cpp\nstd::vector\u003cfloat\u003e input = {...};\nmodel-\u003epredict(input.data());\n```\n\nIf you want to use more complex forward pass, using `Concat` or `Add` layers, you can subclass the model class and override the default `predict` function\n\n```cpp\nclass MyModel : public CUDANet::Model {\n    ...\n}\n\n...\n\nfloat* MyModel::predict(const float* input) {\n    float* d_input = inputLayer-\u003eforward(input);\n\n    d_conv1 = getLayer(\"conv1\")-\u003eforward(d_input);\n    d_conv2 = getLayer(\"conv2\")-\u003eforward(d_input);\n\n    d_output = concatLayer-\u003eforward(d_conv1, d_conv2);\n\n    return outputLayer-\u003eforward(d_input);\n}\n```\n\n### Load Pre-trained Weights\n\nCUDANet uses format similar to safetensors to load weights and biases.\n\n```\n[u_short version, u_int64 header size, header, tensor values]\n```\n\nwhere `header` is a csv format\n\n```\n\u003ctensor_name\u003e,\u003ctensor_size\u003e,\u003ctensor_offset\u003e\n```\n\nTo load weights call `load_weights` function on Model object. To export weights from pytorch you can use the `export_model_weights` function from `tools/utils.py`  script. Currently only float32 weights are supported","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flordmathis%2Fcudanet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flordmathis%2Fcudanet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flordmathis%2Fcudanet/lists"}