{"id":19998145,"url":"https://github.com/iacolippo/gpu-dnn-install","last_synced_at":"2025-08-23T10:36:34.819Z","repository":{"id":94984815,"uuid":"83479389","full_name":"iacolippo/gpu-dnn-install","owner":"iacolippo","description":"Scripts and instructions to install CUDA, cuDNN and the most common deep learning frameworks.","archived":false,"fork":false,"pushed_at":"2017-05-14T11:48:20.000Z","size":92,"stargazers_count":4,"open_issues_count":0,"forks_count":2,"subscribers_count":2,"default_branch":"master","last_synced_at":"2025-05-04T14:46:55.370Z","etag":null,"topics":["cuda-toolkit","cudnn","install-script","theano","torch"],"latest_commit_sha":null,"homepage":null,"language":"Shell","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/iacolippo.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"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,"zenodo":null}},"created_at":"2017-02-28T21:09:09.000Z","updated_at":"2020-02-13T19:06:27.000Z","dependencies_parsed_at":"2023-06-11T10:45:37.762Z","dependency_job_id":null,"html_url":"https://github.com/iacolippo/gpu-dnn-install","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/iacolippo/gpu-dnn-install","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iacolippo%2Fgpu-dnn-install","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iacolippo%2Fgpu-dnn-install/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iacolippo%2Fgpu-dnn-install/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iacolippo%2Fgpu-dnn-install/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/iacolippo","download_url":"https://codeload.github.com/iacolippo/gpu-dnn-install/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iacolippo%2Fgpu-dnn-install/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":271746660,"owners_count":24813575,"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":"2025-08-23T02:00:09.327Z","response_time":69,"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":["cuda-toolkit","cudnn","install-script","theano","torch"],"created_at":"2024-11-13T05:07:29.314Z","updated_at":"2025-08-23T10:36:34.813Z","avatar_url":"https://github.com/iacolippo.png","language":"Shell","funding_links":[],"categories":[],"sub_categories":[],"readme":"# gpu-dnn-install\nAuthor: Iacopo Poli\n\nDescription: Scripts and instructions to install CUDA, cuDNN and two of the most common deep learning frameworks ([Theano](http://deeplearning.net/software/theano/) and [Torch](http://torch.ch)).\n\n## PREREQUISITES:\n\n1 - Download CUDA 8.0 deb(network) file for your system [here](https://developer.nvidia.com/cuda-downloads). If you're using Ubuntu 16.04 on NV6, the file should be called\n\n```bash\ncuda-repo-ubuntu1604_8.0.61-1_amd64.deb\n```\n\n2 - Download CuDNN 5.1 for CUDA 8.0 Linux [here](https://developer.nvidia.com/rdp/cudnn-download). You have to register first and accept the License. The file should be called\n        \n```bash\ncudnn-8.0-linux-x64-v5.1.tar\n```\n\nIt should work the same with new versions of CuDNN.\n\nAll the other files needed are in this repository.\n\n## INSTALLATION\n\nNOTE: You have to set the permission to execute the installation script files. You can do that with \n\n```bash \nchmod a+x \u003cfilename\u003e\n```\n\n0 - Run this and check that it prints something, otherwise there is no NVIDIA hardware available.\n\n```bash\nlspci | grep -i nvidia\n```\n\nSample output:\n\n![alt text](Img/NVIDIA-hardware.png \"Sample output lspci\")\n\n   \n1 - Run \n\n```bash\n./cuda-install.sh\n```\n\n2 - add /usr/local/cuda-8.0/bin to PATH environment variable in .profile in home directory using nano or vim\n\n```bash\nexport PATH=\"$PATH:/usr/local/cuda-8.0/bin\"\n```\n\n3 - add /usr/local/cuda-8.0/lib64\\ to LD_LIBRARY_PATH environment variable in .profile using nano or vim\n   \n```bash\nexport LD_LIBRARY_PATH=\"$LD_LIBRARY_PATH:/usr/local/cuda-8.0/lib64\"\n```\n\nFor points 2 \u0026 3 you can look at the [example file](https://github.com/iacolippo/gpu-dnn-install/blob/master/.profile) in this repository.\n\n4 - Activate the changes using\n\n```bash\nsource .profile\n```\n\n5 - Reboot the system\n        \n```bash\nsudo reboot\n```\n        \n6 - Reconnect to the machine via ssh and write a ```.theanorc``` file in the home directory equal to [this](https://github.com/iacolippo/gpu-dnn-install/blob/master/.theanorc). Then run the following command and check that is using gpu. It should also print a message that cuDNN is not available.\n        \n```bash\n./theano-install.sh\n```\n\nOutput:\n\n![alt text](Img/using-gpu.png \"Sample output gpu usage\")\n\n7 - Run the following command.\n        \n```bash\n./cudnn-install.sh\n```\n\nIf you installed Theano, you can run ```python gpu-test.py``` and you should see cuDNN is now available.\nOutput:\n\n![alt text](Img/using-cudnn.png \"Sample output using cudnn and gpu\")\n\n        \n8 - Install Torch\n\n```bash\n./torch-install.sh\n```\n\nAnswer *yes* to anything on the terminal. At the end, enter\n\n```bash\nsource ~/.bashrc\n```\n\n9 - Install Tensorflow (GPU version)\n\n```bash\n./tensorflow-install.sh\n```\n\nCheck that the GPU is being used by running\n\n```bash\npython tensorflow-gpu.py\n```\n\n10 - Install Keras by running \n\n```bash\nsudo pip install keras\n```\n\nWhen using Tensorflow backend (default setting), the code runs on GPU automatically if one is detected.\n\nFor any question you can contact me on Twitter [@iacopo_poli](https://twitter.com/iacopo_poli).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fiacolippo%2Fgpu-dnn-install","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fiacolippo%2Fgpu-dnn-install","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fiacolippo%2Fgpu-dnn-install/lists"}