{
  "id": 304219,
  "title": "Is it OK to use notebook as dataset during inference?",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/304219",
  "author_name": "",
  "post_date": "2022-01-31T08:53:25.160235200Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Kaggle now allows to add notebook output as input data for other notebook withour creation of a dataset. Is it OK to use this during inference? Or it's better to create dataset, and use it as it was done earlier?</p>",
  "messages": [
    {
      "id": "1670141",
      "postDate": "01/31/2022 08:53:25",
      "content": "<p>Kaggle now allows to add notebook output as input data for other notebook withour creation of a dataset. Is it OK to use this during inference? Or it's better to create dataset, and use it as it was done earlier?</p>",
      "rawMarkdown": "Kaggle now allows to add notebook output as input data for other notebook withour creation of a dataset. Is it OK to use this during inference? Or it's better to create dataset, and use it as it was done earlier?",
      "votes": null
    },
    {
      "id": "1670188",
      "postDate": "01/31/2022 09:54:11",
      "content": "<p>Hi, I am currently using this with some deep learning models ( the model itself  is in the  output of another notebook) in another competition and I was able to run offline and score without any problems so I think its fine ! </p>",
      "rawMarkdown": "Hi, I am currently using this with some deep learning models ( the model itself  is in the  output of another notebook) in another competition and I was able to run offline and score without any problems so I think its fine !",
      "votes": null
    },
    {
      "id": "1670238",
      "postDate": "01/31/2022 10:48:36",
      "content": "<p>I see, thanks for your answer</p>",
      "rawMarkdown": "I see, thanks for your answer",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1670188,
      "author_name": "asalhi",
      "author_url": "",
      "post_date": "01/31/2022 09:54:11",
      "content": "<p>Hi, I am currently using this with some deep learning models ( the model itself  is in the  output of another notebook) in another competition and I was able to run offline and score without any problems so I think its fine ! </p>",
      "votes": null,
      "replies": [
        {
          "id": 1670238,
          "author_name": "atamazian",
          "author_url": "",
          "post_date": "01/31/2022 10:48:36",
          "content": "<p>I see, thanks for your answer</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1670141": "Kaggle now allows to add notebook output as input data for other notebook withour creation of a dataset. Is it OK to use this during inference? Or it's better to create dataset, and use it as it was done earlier?",
    "1670188": "Hi, I am currently using this with some deep learning models ( the model itself  is in the  output of another notebook) in another competition and I was able to run offline and score without any problems so I think its fine !",
    "1670238": "I see, thanks for your answer"
  },
  "source": "meta"
}