{
  "id": 321922,
  "title": "What is your setup/stack for this competition?",
  "url": "/competitions/h-and-m-personalized-fashion-recommendations/discussion/321922",
  "author_name": "",
  "post_date": "2022-04-29T11:07:15.622891800Z",
  "votes": 2,
  "comment_count": 7,
  "views": 0,
  "content": "<p>Until now, I mainly used kaggle notebooks to participate to competitions but I feel my options are getting limited. I was wondering out of curiosity what type of setups you were using on your side to tackle competitions efficiently? Are most of you running code on notebooks or do you have a more elaborate setup?</p>\n<p><strong>A few topics I can think of:</strong><br>\nProcessing: Kaggle/Local/Google Cloud…?<br>\nCoding: Notebook/IDE (like Pycharm)…?<br>\nModules: main file running different modules with a config file/Apache airflow…?<br>\nParameters tuning tracking: In-house/MLFlow…?<br>\nVersion control: Git/nothing…</p>\n<p>Anything I forgot?</p>\n<p>And if you are working in teams, how do you split the tasks?</p>\n<p>Thanks!</p>",
  "messages": [
    {
      "id": "1771566",
      "postDate": "04/29/2022 11:07:15",
      "content": "<p>Until now, I mainly used kaggle notebooks to participate to competitions but I feel my options are getting limited. I was wondering out of curiosity what type of setups you were using on your side to tackle competitions efficiently? Are most of you running code on notebooks or do you have a more elaborate setup?</p>\n<p><strong>A few topics I can think of:</strong><br>\nProcessing: Kaggle/Local/Google Cloud…?<br>\nCoding: Notebook/IDE (like Pycharm)…?<br>\nModules: main file running different modules with a config file/Apache airflow…?<br>\nParameters tuning tracking: In-house/MLFlow…?<br>\nVersion control: Git/nothing…</p>\n<p>Anything I forgot?</p>\n<p>And if you are working in teams, how do you split the tasks?</p>\n<p>Thanks!</p>",
      "rawMarkdown": "Until now, I mainly used kaggle notebooks to participate to competitions but I feel my options are getting limited. I was wondering out of curiosity what type of setups you were using on your side to tackle competitions efficiently? Are most of you running code on notebooks or do you have a more elaborate setup?\n\n**A few topics I can think of:**\nProcessing: Kaggle/Local/Google Cloud...?\nCoding: Notebook/IDE (like Pycharm)...?\nModules: main file running different modules with a config file/Apache airflow...?\nParameters tuning tracking: In-house/MLFlow...?\nVersion control: Git/nothing...\n\nAnything I forgot?\n\nAnd if you are working in teams, how do you split the tasks?\n\nThanks!",
      "votes": null
    },
    {
      "id": "1771820",
      "postDate": "04/29/2022 15:53:16",
      "content": "<p>I'm pushing my local git repo files with all my helper functions to a kaggle dataset.<br>\nThat way I can have version control on most of my code, and the notebooks are much cleaner and manageable.</p>",
      "rawMarkdown": "I'm pushing my local git repo files with all my helper functions to a kaggle dataset.\nThat way I can have version control on most of my code, and the notebooks are much cleaner and manageable.",
      "votes": null
    },
    {
      "id": "1771901",
      "postDate": "04/29/2022 16:38:53",
      "content": "<p>Right! Makes sense! Thanks!</p>",
      "rawMarkdown": "Right! Makes sense! Thanks!",
      "votes": null
    },
    {
      "id": "1771946",
      "postDate": "04/29/2022 17:06:19",
      "content": "<p>Another tool that seems very powerful - Kedro: <a href=\"https://github.com/adrian-dembek/kaggle-hm-kedro\" target=\"_blank\">https://github.com/adrian-dembek/kaggle-hm-kedro</a></p>",
      "rawMarkdown": "Another tool that seems very powerful - Kedro: https://github.com/adrian-dembek/kaggle-hm-kedro",
      "votes": null
    },
    {
      "id": "1773649",
      "postDate": "05/01/2022 10:56:45",
      "content": "<p>I'm using an IDE with a Jupyter plugin (PyCharm), and I have a local git repository with all my helper functions. I also use a parameters tuning tracking tool (MLFlow) and a code version control system (Git).</p>",
      "rawMarkdown": "I'm using an IDE with a Jupyter plugin (PyCharm), and I have a local git repository with all my helper functions. I also use a parameters tuning tracking tool (MLFlow) and a code version control system (Git).",
      "votes": null
    },
    {
      "id": "1775694",
      "postDate": "05/03/2022 09:25:16",
      "content": "<p>Are you running your code on your local machine then?</p>",
      "rawMarkdown": "Are you running your code on your local machine then?",
      "votes": null
    },
    {
      "id": "1776371",
      "postDate": "05/04/2022 00:19:21",
      "content": "<p>MacBook pro: coding in the pycharm;<br>\nA Linux machine with large ram: running the code in the terminal, and debugging in the Jupyter notebook;<br>\nGit is used to control code version and to transport the code from MacBook to Linux</p>",
      "rawMarkdown": "MacBook pro: coding in the pycharm;\nA Linux machine with large ram: running the code in the terminal, and debugging in the Jupyter notebook;\nGit is used to control code version and to transport the code from MacBook to Linux",
      "votes": null
    },
    {
      "id": "1777003",
      "postDate": "05/04/2022 10:46:53",
      "content": "<p>Awesome! Thanks!</p>",
      "rawMarkdown": "Awesome! Thanks!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1771820,
      "author_name": "jacob34",
      "author_url": "",
      "post_date": "04/29/2022 15:53:16",
      "content": "<p>I'm pushing my local git repo files with all my helper functions to a kaggle dataset.<br>\nThat way I can have version control on most of my code, and the notebooks are much cleaner and manageable.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1771901,
          "author_name": "loicge",
          "author_url": "",
          "post_date": "04/29/2022 16:38:53",
          "content": "<p>Right! Makes sense! Thanks!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1771946,
      "author_name": "loicge",
      "author_url": "",
      "post_date": "04/29/2022 17:06:19",
      "content": "<p>Another tool that seems very powerful - Kedro: <a href=\"https://github.com/adrian-dembek/kaggle-hm-kedro\" target=\"_blank\">https://github.com/adrian-dembek/kaggle-hm-kedro</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1773649,
      "author_name": "",
      "author_url": "",
      "post_date": "05/01/2022 10:56:45",
      "content": "<p>I'm using an IDE with a Jupyter plugin (PyCharm), and I have a local git repository with all my helper functions. I also use a parameters tuning tracking tool (MLFlow) and a code version control system (Git).</p>",
      "votes": null,
      "replies": [
        {
          "id": 1775694,
          "author_name": "loicge",
          "author_url": "",
          "post_date": "05/03/2022 09:25:16",
          "content": "<p>Are you running your code on your local machine then?</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1776371,
      "author_name": "sirius81",
      "author_url": "",
      "post_date": "05/04/2022 00:19:21",
      "content": "<p>MacBook pro: coding in the pycharm;<br>\nA Linux machine with large ram: running the code in the terminal, and debugging in the Jupyter notebook;<br>\nGit is used to control code version and to transport the code from MacBook to Linux</p>",
      "votes": null,
      "replies": [
        {
          "id": 1777003,
          "author_name": "loicge",
          "author_url": "",
          "post_date": "05/04/2022 10:46:53",
          "content": "<p>Awesome! Thanks!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1771566": "Until now, I mainly used kaggle notebooks to participate to competitions but I feel my options are getting limited. I was wondering out of curiosity what type of setups you were using on your side to tackle competitions efficiently? Are most of you running code on notebooks or do you have a more elaborate setup?\n\n**A few topics I can think of:**\nProcessing: Kaggle/Local/Google Cloud...?\nCoding: Notebook/IDE (like Pycharm)...?\nModules: main file running different modules with a config file/Apache airflow...?\nParameters tuning tracking: In-house/MLFlow...?\nVersion control: Git/nothing...\n\nAnything I forgot?\n\nAnd if you are working in teams, how do you split the tasks?\n\nThanks!",
    "1771820": "I'm pushing my local git repo files with all my helper functions to a kaggle dataset.\nThat way I can have version control on most of my code, and the notebooks are much cleaner and manageable.",
    "1771901": "Right! Makes sense! Thanks!",
    "1771946": "Another tool that seems very powerful - Kedro: https://github.com/adrian-dembek/kaggle-hm-kedro",
    "1773649": "I'm using an IDE with a Jupyter plugin (PyCharm), and I have a local git repository with all my helper functions. I also use a parameters tuning tracking tool (MLFlow) and a code version control system (Git).",
    "1775694": "Are you running your code on your local machine then?",
    "1776371": "MacBook pro: coding in the pycharm;\nA Linux machine with large ram: running the code in the terminal, and debugging in the Jupyter notebook;\nGit is used to control code version and to transport the code from MacBook to Linux",
    "1777003": "Awesome! Thanks!"
  },
  "source": "meta"
}