{
  "id": 488385,
  "title": "[need kaggle staff help] Can I Create my Dockerfile for submission?",
  "url": "/competitions/hms-harmful-brain-activity-classification/discussion/488385",
  "author_name": "",
  "post_date": "2024-04-02T08:57:31.246731500Z",
  "votes": null,
  "comment_count": 5,
  "views": 0,
  "content": "<p>It takes me about 30mins to configure the environment in the notebook (installation package, etc.). Is there a way for me to customize my Dockerfile so that I don't spend too much time on the environment?</p>",
  "messages": [
    {
      "id": "2728472",
      "postDate": "04/02/2024 08:57:31",
      "content": "<p>It takes me about 30mins to configure the environment in the notebook (installation package, etc.). Is there a way for me to customize my Dockerfile so that I don't spend too much time on the environment?</p>",
      "rawMarkdown": "It takes me about 30mins to configure the environment in the notebook (installation package, etc.). Is there a way for me to customize my Dockerfile so that I don't spend too much time on the environment?",
      "votes": null
    },
    {
      "id": "2729307",
      "postDate": "04/02/2024 16:25:37",
      "content": "<p>There isn't. With that said, once the competition is over could you share a relevant notebook?</p>",
      "rawMarkdown": "There isn't. With that said, once the competition is over could you share a relevant notebook?",
      "votes": null
    },
    {
      "id": "2729328",
      "postDate": "04/02/2024 16:39:26",
      "content": "<p>i work on kaggle's docker image, im interested in seeing what configuration you are doing as well. please share with me too.</p>",
      "rawMarkdown": "i work on kaggle's docker image, im interested in seeing what configuration you are doing as well. please share with me too.",
      "votes": null
    },
    {
      "id": "2729341",
      "postDate": "04/02/2024 16:48:26",
      "content": "<p><a href=\"https://github.com/Kaggle/docker-python/tree/main\" target=\"_blank\">https://github.com/Kaggle/docker-python/tree/main</a></p>",
      "rawMarkdown": "https://github.com/Kaggle/docker-python/tree/main",
      "votes": null
    },
    {
      "id": "2732099",
      "postDate": "04/03/2024 02:24:07",
      "content": "<blockquote>\n  <p>There isn't. With that said, once the competition is over could you share a relevant notebook?</p>\n</blockquote>\n<p>Thank you very much, I was actually planning to use mamba_ssm. since I need to compile every time I configure the environment, it takes about 30 minutes to configure the environment. Now that I have compiled a whl file, it only takes 2 minute to configure the environment again.</p>\n<p>Before: <a href=\"https://www.kaggle.com/code/jimmyisme1/mamba-build/notebook\" target=\"_blank\">https://www.kaggle.com/code/jimmyisme1/mamba-build/notebook</a><br>\nAfter: <a href=\"https://www.kaggle.com/code/jimmyisme1/test-lib/notebook\" target=\"_blank\">https://www.kaggle.com/code/jimmyisme1/test-lib/notebook</a></p>",
      "rawMarkdown": "> There isn't. With that said, once the competition is over could you share a relevant notebook?\n\nThank you very much, I was actually planning to use mamba_ssm. since I need to compile every time I configure the environment, it takes about 30 minutes to configure the environment. Now that I have compiled a whl file, it only takes 2 minute to configure the environment again.\n\nBefore: https://www.kaggle.com/code/jimmyisme1/mamba-build/notebook\nAfter: https://www.kaggle.com/code/jimmyisme1/test-lib/notebook",
      "votes": null
    },
    {
      "id": "2733642",
      "postDate": "04/03/2024 18:22:20",
      "content": "<p>Storing wheel files into a dataset is a common practice done by kagglers to bring in packages that not in our docker image for competition use. we try to offer a diverse selection of package, but some package are too niche or not as popular to be added to our image since it can cause mainitance issues on our end.</p>\n<p>i recommend taking a peak at these notebooks that use this approach, may help with other notebooks.</p>\n<p><a href=\"https://www.kaggle.com/code/samuelepino/pip-downloading-packages-to-your-local-machine\" target=\"_blank\">https://www.kaggle.com/code/samuelepino/pip-downloading-packages-to-your-local-machine</a><br>\n<a href=\"https://www.kaggle.com/code/samuelepino/pip-installing-packages-with-no-internet\" target=\"_blank\">https://www.kaggle.com/code/samuelepino/pip-installing-packages-with-no-internet</a> </p>",
      "rawMarkdown": "Storing wheel files into a dataset is a common practice done by kagglers to bring in packages that not in our docker image for competition use. we try to offer a diverse selection of package, but some package are too niche or not as popular to be added to our image since it can cause mainitance issues on our end.\n\ni recommend taking a peak at these notebooks that use this approach, may help with other notebooks.\n\nhttps://www.kaggle.com/code/samuelepino/pip-downloading-packages-to-your-local-machine\nhttps://www.kaggle.com/code/samuelepino/pip-installing-packages-with-no-internet",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2729307,
      "author_name": "sohier",
      "author_url": "",
      "post_date": "04/02/2024 16:25:37",
      "content": "<p>There isn't. With that said, once the competition is over could you share a relevant notebook?</p>",
      "votes": null,
      "replies": [
        {
          "id": 2729328,
          "author_name": "jcchavez",
          "author_url": "",
          "post_date": "04/02/2024 16:39:26",
          "content": "<p>i work on kaggle's docker image, im interested in seeing what configuration you are doing as well. please share with me too.</p>",
          "votes": null,
          "replies": [
            {
              "id": 2729341,
              "author_name": "cdeotte",
              "author_url": "",
              "post_date": "04/02/2024 16:48:26",
              "content": "<p><a href=\"https://github.com/Kaggle/docker-python/tree/main\" target=\"_blank\">https://github.com/Kaggle/docker-python/tree/main</a></p>",
              "votes": null,
              "replies": []
            }
          ]
        },
        {
          "id": 2732099,
          "author_name": "jimmyisme1",
          "author_url": "",
          "post_date": "04/03/2024 02:24:07",
          "content": "<blockquote>\n  <p>There isn't. With that said, once the competition is over could you share a relevant notebook?</p>\n</blockquote>\n<p>Thank you very much, I was actually planning to use mamba_ssm. since I need to compile every time I configure the environment, it takes about 30 minutes to configure the environment. Now that I have compiled a whl file, it only takes 2 minute to configure the environment again.</p>\n<p>Before: <a href=\"https://www.kaggle.com/code/jimmyisme1/mamba-build/notebook\" target=\"_blank\">https://www.kaggle.com/code/jimmyisme1/mamba-build/notebook</a><br>\nAfter: <a href=\"https://www.kaggle.com/code/jimmyisme1/test-lib/notebook\" target=\"_blank\">https://www.kaggle.com/code/jimmyisme1/test-lib/notebook</a></p>",
          "votes": null,
          "replies": [
            {
              "id": 2733642,
              "author_name": "jcchavez",
              "author_url": "",
              "post_date": "04/03/2024 18:22:20",
              "content": "<p>Storing wheel files into a dataset is a common practice done by kagglers to bring in packages that not in our docker image for competition use. we try to offer a diverse selection of package, but some package are too niche or not as popular to be added to our image since it can cause mainitance issues on our end.</p>\n<p>i recommend taking a peak at these notebooks that use this approach, may help with other notebooks.</p>\n<p><a href=\"https://www.kaggle.com/code/samuelepino/pip-downloading-packages-to-your-local-machine\" target=\"_blank\">https://www.kaggle.com/code/samuelepino/pip-downloading-packages-to-your-local-machine</a><br>\n<a href=\"https://www.kaggle.com/code/samuelepino/pip-installing-packages-with-no-internet\" target=\"_blank\">https://www.kaggle.com/code/samuelepino/pip-installing-packages-with-no-internet</a> </p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2728472": "It takes me about 30mins to configure the environment in the notebook (installation package, etc.). Is there a way for me to customize my Dockerfile so that I don't spend too much time on the environment?",
    "2729307": "There isn't. With that said, once the competition is over could you share a relevant notebook?",
    "2729328": "i work on kaggle's docker image, im interested in seeing what configuration you are doing as well. please share with me too.",
    "2729341": "https://github.com/Kaggle/docker-python/tree/main",
    "2732099": "> There isn't. With that said, once the competition is over could you share a relevant notebook?\n\nThank you very much, I was actually planning to use mamba_ssm. since I need to compile every time I configure the environment, it takes about 30 minutes to configure the environment. Now that I have compiled a whl file, it only takes 2 minute to configure the environment again.\n\nBefore: https://www.kaggle.com/code/jimmyisme1/mamba-build/notebook\nAfter: https://www.kaggle.com/code/jimmyisme1/test-lib/notebook",
    "2733642": "Storing wheel files into a dataset is a common practice done by kagglers to bring in packages that not in our docker image for competition use. we try to offer a diverse selection of package, but some package are too niche or not as popular to be added to our image since it can cause mainitance issues on our end.\n\ni recommend taking a peak at these notebooks that use this approach, may help with other notebooks.\n\nhttps://www.kaggle.com/code/samuelepino/pip-downloading-packages-to-your-local-machine\nhttps://www.kaggle.com/code/samuelepino/pip-installing-packages-with-no-internet"
  },
  "source": "meta"
}