{"id":19165584,"url":"https://github.com/lablup/backend.ai-kernels","last_synced_at":"2026-03-07T18:01:21.689Z","repository":{"id":12792898,"uuid":"44731840","full_name":"lablup/backend.ai-kernels","owner":"lablup","description":"Repository of Backend.AI-enabled container recipes","archived":false,"fork":false,"pushed_at":"2025-04-14T06:50:49.000Z","size":317335,"stargazers_count":33,"open_issues_count":92,"forks_count":14,"subscribers_count":17,"default_branch":"main","last_synced_at":"2025-04-15T03:54:41.933Z","etag":null,"topics":["alpine-linux","deep-learning","docker","programming-languages","repl","sandbox","sorna","ubuntu"],"latest_commit_sha":null,"homepage":"https://www.backend.ai","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"lgpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/lablup.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,"zenodo":null}},"created_at":"2015-10-22T08:22:13.000Z","updated_at":"2025-04-14T06:50:54.000Z","dependencies_parsed_at":"2023-10-21T15:35:15.053Z","dependency_job_id":"61d6f61d-d434-42ce-b874-8200f48f746a","html_url":"https://github.com/lablup/backend.ai-kernels","commit_stats":null,"previous_names":[],"tags_count":1,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lablup%2Fbackend.ai-kernels","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lablup%2Fbackend.ai-kernels/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lablup%2Fbackend.ai-kernels/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lablup%2Fbackend.ai-kernels/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/lablup","download_url":"https://codeload.github.com/lablup/backend.ai-kernels/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":249003954,"owners_count":21196794,"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","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":["alpine-linux","deep-learning","docker","programming-languages","repl","sandbox","sorna","ubuntu"],"created_at":"2024-11-09T09:28:32.904Z","updated_at":"2026-03-07T18:01:16.648Z","avatar_url":"https://github.com/lablup.png","language":"Jupyter Notebook","readme":"# Backend.AI Kernels Repository\n\nBackend.AI agent kernels in various programming languages / toolkits and frameworks.\n\n\n## Officially Supported Images\n\n### Warrented Images (Deep Learning Framework)\n\nHere we list the latest versions of our supported kernel images.\n\"\\*\" in the Query mode column means that it supports preservation of global contexts across different query runs.\n\n| Deep-Learning Framework | Image Name                 | Version | Batch | Query | Input Hook | TTY | Runtime Impl.     |\n|-------------------------|----------------------------|---------|-------|-------|------------|-----|-------------------|\n| TensorFlow              | `lablup/python-tensorflow` | 1.15.4  | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.4.3   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.5.2   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.6.0   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.7.0   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.8.0   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.9.0   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.10.0  | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.11.0  | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.12.0  | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.13.0  | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.14.0  | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| TensorFlow              | `lablup/python-tensorflow` | 2.15.0  | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| PyTorch                 | `lablup/python-torch`      | 1.7     | O     | O\\*   | O          |     |                   |\n| PyTorch                 | `lablup/python-torch`      | 1.8     | O     | O\\*   | O          |     |                   |\n| PyTorch                 | `lablup/python-torch`      | 1.9     | O     | O\\*   | O          |     |                   |\n| PyTorch                 | `lablup/python-torch`      | 1.10    | O     | O\\*   | O          |     |                   |\n| PyTorch                 | `lablup/python-torch`      | 1.11    | O     | O\\*   | O          |     |                   |\n| PyTorch                 | `lablup/python-torch`      | 1.12    | O     | O\\*   | O          |     |                   |\n| PyTorch                 | `lablup/python-torch`      | 2.0     | O     | O\\*   | O          |     |                   |\n| PyTorch                 | `lablup/python-torch`      | 2.1     | O     | O\\*   | O          |     |                   |\n| CNTK                    | `lablup/python-cntk`       | 2.6     | O     | O\\*   | O          |     |                   |\n| CNTK                    | `lablup/python-cntk`       | 2.7     | O     | O\\*   | O          |     |                   |\n| MXnet                   | `lablup/python-mxnet`      | 1.4.1   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| MXnet                   | `lablup/python-mxnet`      | 1.5.1   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| MXnet                   | `lablup/python-mxnet`      | 1.6.0   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| MXnet                   | `lablup/python-mxnet`      | 1.7.0   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| MXnet                   | `lablup/python-mxnet`      | 1.8.0   | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| All-in-one Environment  | `lablup/python-ff`         | 20.09   | O     | O\\*   | O          |     |                   |\n| All-in-one Environment  | `lablup/python-ff`         | 21.01   | O     | O\\*   | O          |     |                   |\n| All-in-one Environment  | `lablup/python-ff`         | 21.03   | O     | O\\*   | O          |     |                   |\n| All-in-one Environment  | `lablup/python-ff`         | 21.08   | O     | O\\*   | O          |     |                   |\n| All-in-one Environment  | `lablup/python-ff`         | 21.09   | O     | O\\*   | O          |     |                   |\n| All-in-one Environment  | `lablup/python-ff`         | 22.02   | O     | O\\*   | O          |     |                   |\n\n### Warrented Images (NGC, Nvidia GPU Cloud)\n| Deep-Learning Framework | Image Name              | Version       | Batch | Query | Input Hook | TTY | Runtime Impl.     |\n|-------------------------|-------------------------|---------------|-------|-------|------------|-----|-------------------|\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.07-py2     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.07-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.08-py2     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.08-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.09-py2     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.09-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.10-py2     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.10-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.11-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.11-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.12-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 19.12-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.01-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.01-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.02-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.02-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.03-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.03-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.06-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.06-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.07-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.07-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.08-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.08-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.09-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.09-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.10-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.10-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.11-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.11-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.12-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 20.12-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.03-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.03-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.05-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.05-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.06-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.06-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.07-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.07-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.08-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.08-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.09-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.09-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.10-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.10-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.11-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.11-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.12-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 21.12-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.01-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.01-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.01-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.02-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.02-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.03-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.03-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.04-tf2-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-TensorFlow          | `lablup/ngc-tensorflow` | 22.04-tf1-py3 | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 19.07-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 19.08-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 19.09-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 19.10-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 19.11-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 19.12-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.01-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.02-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.03-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.06-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.07-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.08-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.09-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.10-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.11-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 20.12-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.03-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.04-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.05-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.06-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.07-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.08-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.09-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 21.11-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 22.02-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 23.03-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 23.07-py3     | O     | O\\*   | O          |     |                   |\n| NGC-Pytorch             | `lablup/ngc-pytorch`    | 23.12-py3     | O     | O\\*   | O          |     |                   |\n| NGC-MXNet               | `lablup/ngc-mxnet`      | 19.07-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-MXNet               | `lablup/ngc-mxnet`      | 19.08-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-MXNet               | `lablup/ngc-mxnet`      | 19.09-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-MXNet               | `lablup/ngc-mxnet`      | 19.10-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-MXNet               | `lablup/ngc-mxnet`      | 19.11-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-MXNet               | `lablup/ngc-mxnet`      | 19.12-py3     | O     | O\\*   | O          |     | Bundled w/Keras 2 |\n| NGC-MATLAB              | `lablup/ngc-matlab`     | 2019b\t    | O     | O\\*   | O          |     | \t \t   |\n\n### Commnunity-maintained Languages\n\n| Language      | Image Name              | Version         | Batch | Query | Input Hook | TTY | Runtime Impl. |\n|---------------|-------------------------|-----------------|-------|-------|---|---|--------------------|\n| C             | `lablup/kernel-c`       | 6.3             | O     | O     | O |   | GCC on Alpine 3.8  |\n| C++ (14)      | `lablup/kernel-cpp`     | 6.3             | O     | O     |   |   | GCC on Alpine 3.8  |\n| Go            | `lablup/kernel-go`      | 1.9             | O     | O     |   |   |                    |\n| Haskell       | `lablup/kernel-haskell` | 8.2             | O     | O     |   |   |                    |\n| Java          | `lablup/kernel-java`    | 8.0             | O     | O     |   |   |                    |\n| Linux Console | `lablup/kernel-git`     | -               | -     | -     | - | O | Bash on Alpine 3.8 |\n| Lua           | `lablup/kernel-lua`     | 5.3             | O     | O     |   |   |                    |\n| Node.js       | `lablup/kernel-nodejs`  | 6.14,8.11,10.11 | O     | O     |   |   |                    |\n| Octave        | `lablup/kernel-octave`  | 4.2             | O     | O     |   |   |                    |\n| Python        | `lablup/kernel-python`  | 2.7             | O     | O     | O |   | beta               |\n| Python        | `lablup/kernel-python`  | 3.6.6           | O     | O\\*   | O |   |                    |\n| Rust          | `lablup/kernel-rust`    | 1.17            | O     | O     |   |   |                    |\n| PHP           | `lablup/kernel-php`     | 7.1             | O     | O     |   |   |                    |\n| R             | `lablup/kernel-r`       | 3.3             | O     | O     |   |   | CRAN R             |\n| Scala\t        | `lablup/kernel-scala`   | 2.12            | O     | O     |   |   |                    |\n\n\n### Commnunity-maintained \u0026 Deep learning based images\n * `base-mkl`    (Intel' Machine Learning Kits (MKL) works on CPU only kernel)\n * `base-cuda`   (Nvidia' GPU \u0026 CUDA libarary compatibility, needed Nvidia-docker)\n * `base-TPU`    (Google TPU comptibility, on Google' Cloud)\n * `base-ROCm`   (AMD' GPU \u0026 OpenCL libarary compatibility, T.B.D.)\n \n| base-mkl (Intel)      | base-cuda (Nvidia)             | base-TPU (Google)        | base-ROCm (AMD)           |\n|-----------------------|--------------------------------|--------------------------|---------------------------|\n| tensorflow-2.3-py36   | tensorflow-2.3-py38-cuda10     |\t                \t    |                           |\n| tensorflow-2.2-py36   | tensorflow-2.2-py36-cuda10     |                          |                           |\n| tensorflow-2.1-py36   | tensorflow-2.1-py36-cuda10     |\t                \t    |                           |\n| tensorflow-2.0-py36   | tensorflow-2.0-py36-cuda10     |                          |                           |\n| tensorflow-1.15-py36  | tensorflow-1.15-py36-cuda10    |\t                \t    |                           |\n| tensorflow-1.14-py36  | tensorflow-1.14-py36-cuda10    | tensorflow-1.14-py36-tpu | tensorflow-1.14-py36-rocm |\n| tensorflow-1.13-py36  | tensorflow-1.13-py36-cuda10    | tensorflow-1.13-py36-tpu |                           |\n| tensorflow-1.12-py36  | tensorflow-1.12-py36-cuda9     |                          |                           |\n| tensorflow-1.11-py36  | tensorflow-1.11-py36-cuda9     |                          |                           |\n| tensorflow-1.10-py36  | tensorflow-1.10-py36-cuda9     |                          |                           |\n| tensorflow-1.9-py36   | tensorflow-1.9-py36-cuda9      |                          |                           |\n| tensorflow-1.8-py36   | tensorflow-1.8-py36-cuda9      |                          |                           |\n| tensorflow-1.7-py36   | tensorflow-1.7-py36-cuda9      |                          |                           |\n| tensorflow-1.6-py36   | tensorflow-1.6-py36-cuda9      |                          |                           |\n| tensorflow-1.5-py36   | tensorflow-1.5-py36-cuda9      |                          |                           |\n| tensorflow-1.4-py36   | tensorflow-1.4-py36-cuda8      |                          |                           |\n| tensorflow-1.3-py36   | tensorflow-1.3-py36-cuda8      |                          |                           |\n| tensorflow-1.2-py36   | tensorflow-1.2-py36-cuda8      |                          |                           |\n| tensorflow-1.1-py36   | tensorflow-1.1-py36-cuda8      |                          |                           |\n| tensorflow-1.0-py36   | tensorflow-1.0-py36-cuda8      |                          |                           |\n|                       | python-caffe2-1.0-py36-cuda9   |                          |                           |\n|                       | python-torch-1.5-py36-cuda10.1 |                          |                           |\n|                       | python-torch-1.4-py36-cuda10.1 |                          |                           |\n|                       | python-torch-1.3-py36-cuda10.1 |                          |                           |\n|                       | python-torch-1.2-py36-cuda9    |                          |                           |\n|                       | python-torch-1.1-py36-cuda9    |                          |                           |\n| python-torch-1.0-py36 | python-torch-1.0-py36-cuda9    |                          |                           |\n| python-torch-0.4-py36 | python-torch-0.4-py36-cuda9    |                          |                           |\n| python-torch-0.3-py36 | python-torch-0.3-py36-cuda9    |                          |                           |\n| python-torch-0.2-py36 | python-torch-0.2-py36-cuda8    |                          |                           |\n|                       | python-mxnet-1.5-py36-cuda10.1 |                          |                           |\n|                       | python-mxnet-1.4-py36-cuda10 |                          |                           |\n|                       | python-mxnet-1.3-py36-cuda10 |                          |                           |\n|                       | python-mxnet-1.2-py36-cuda9  |                          |                           |\n|                       | python-mxnet-1.1-py36-cuda9  |                          |                           |\n|                       | python-mxnet-1.0-py36-cuda9  |                          |                           |\n| python-cntk-2.7-py36  | python-cntk-2.7-py36-cuda9   |                          |                           |\n| python-cntk-2.6-py36  | python-cntk-2.6-py36-cuda9   |                          |                           |\n| python-cntk-2.5-py36  | python-cntk-2.5-py36-cuda9   |                          |                           |\n| python-cntk-2.4-py36  | python-cntk-2.4-py36-cuda9   |                          |                           |\n| python-cntk-2.3-py36  | python-cntk-2.3-py36-cuda9   |                          |                           |\n| python-cntk-2.2-py36  | python-cntk-2.2-py36-cuda9   |                          |                           |\n| python-cntk-2.1-py36  | python-cntk-2.1-py36-cuda9   |                          |                           |\n| python-cntk-2.0-py36  | python-cntk-2.0-py36-cuda9   |                          |                           |\n\n## HOWTO: Adding a New Image\n\nSince Backend.AI v19.03, the kernel-runner component are completely separated from\nthe kernel images as they are mounted at runtime by the agent.\n\nAll you need to do for a new kernel is specifying a set of Backend.AI-specific labels\nand preparation of the jail policy.\n\n* `ai.backend.kernelspec`: For now, it's set to \"1\".\n* `ai.backend.features`: A list of constant strings indicating which Backend.AI\n  kernel features are available for the kernel.\n  - **batch**: Can execute user programs passed as files.\n  - **query**: Can execute user programs passed as code snippets while keeping the\n    context across multiple executions.\n  - **uid-match**: As of 19.03, this must be specified always.\n  - **user-input**: The query/batch mode supports interactive user inputs.\n* `ai.backend.resource.min.*`: The minimum amount of resource to launch this kernel.\n  At least, you must define the CPU core (`cpu`) and the main memory (`mem`).\n  In the memory size values, you may use binary scale-suffixes such as `m` for `MiB`, `g` for `GiB`, etc.\n* `ai.backend.base-distro`: Either \"ubuntu16.04\" or \"alpine3.8\".  Note that Ubuntu\n  18.04-based kernels also need to use \"ubuntu16.04\" here.\n* `ai.backend.runtime-type`: The type of kernel runner to use. (One of the\n  directories in [the `ai.backend.kernel` namespace](https://github.com/lablup/backend.ai-agent/tree/master/src/ai/backend/kernel).)\n* `ai.backend.runtime-path`: The path to the language runtime executable.\n* `ai.backend.service-ports`: A list of 3-tuple strings specifying services available\n  via network tunneling. Each tuple consists of the service name, the service type\n  (one of **pty**, **http**, or **tcp**) and the container-side port number.\n  Backend.AI manages the host-side port mapping and network tunneling via the API\n  gateway automagically.\n* `ai.backend.envs.corecount`: The list of environment variables to be set as the\n  number of available CPU cores to the container.  This is for legacy parallel\n  computation libraries.\n\nNote that the implementation of query/batch modes, runtime-type and service-ports are\nthe responsibility of the kernel runner in the agent codebase.\nFor most computation kernels based Python (e.g., Anaconda, NVIDIA GPU Cloud, etc.)\nmay simply reuse the implementation and labels from the standard \"python\" image.\n\nCurrently we support two major base Linux distros, Ubuntu and Alpine.\n\n### Example: An Ubuntu-based kernel\n\n```dockerfile\nFROM ubuntu:16.04\n\n# Add commands for image customization\nRUN apt-get install ...\n\n# Backend.AI specifics\nCOPY policy.yml /etc/backend.ai/jail/policy.yml\nLABEL ai.backend.kernelspec=1 \\\n      ai.backend.resource.min.cpu=1 \\\n      ai.backend.resource.min.mem=256m \\\n      ai.backend.envs.corecount=\"OPENBLAS_NUM_THREADS,OMP_NUM_THREADS,NPROC\" \\\n      ai.backend.features=\"batch query uid-match user-input\" \\\n      ai.backend.base-distro=\"ubuntu16.04\" \\\n      ai.backend.runtime-type=\"python\" \\\n      ai.backend.runtime-path=\"/usr/local/bin/python\" \\\n      ai.backend.service-ports=\"ipython:pty:3000,jupyter:http:8080\"\n```\n\n### Example: Kernels supporting accelerators\n\nCUDA-accelerated:\n```\n...\nLABEL ... \\\n      ai.backend.resource.min.cuda.device=1 \\\n      ai.backend.resource.min.cuda.smp=2 \\\n      ai.backend.resource.min.cuda.mem=256m \\\n      ...\n...\n```\n\nTPU-accelerated:\n```\n...\nLABEL ... \\\n      ai.backend.resource.min.tpu.device=1 \\\n      ...\n...\n```\n\n### Example: An Alpine-based kernel\n\nAlpine Linux requires two additional lines as it does not support the full-featured\n`ldconfig`.\n\n```dockerfile\nFROM alpine:3.8\n\n# Add commands for image customization\nRUN apk add ...\n\n# Backend.AI specifics\nENV LD_LIBRARY_PATH=/opt/backend.ai/lib\nRUN apk add --no-cache libffi libzmq\nCOPY policy.yml /etc/backend.ai/jail/policy.yml\nLABEL ai.backend.kernelspec=1 \\\n      ai.backend.resource.min.cpu=1 \\\n      ai.backend.resource.min.mem=256m \\\n      ai.backend.features=\"batch query uid-match\" \\\n      ai.backend.base-distro=\"alpine3.8\" \\\n      ai.backend.runtime-type=\"lua\" \\\n      ai.backend.runtime-path=\"/usr/bin/lua\" \\\n      ai.backend.service-ports=\"\"\n```\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flablup%2Fbackend.ai-kernels","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flablup%2Fbackend.ai-kernels","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flablup%2Fbackend.ai-kernels/lists"}