{
  "id": 172989,
  "title": "Are model trained offline legit ?",
  "url": "/competitions/landmark-retrieval-2020/discussion/172989",
  "author_name": "",
  "post_date": "2020-08-07T09:51:33.219564700Z",
  "votes": 8,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Looking at the competition rules I see this \n<code>External data, freely &amp; publicly available, is allowed. This includes pre-trained models.</code></p>\n\n<p>I believe this means training a model offline and uploading it for submission is not allowed.</p>\n\n<p>I remember the organizers said in Bengali : \n<code>External data is allowed, and you are encouraged to train your model offline and use your Notebook for inference.</code></p>\n\n<p>Only models trained inside of a notebook are legit ? </p>\n\n<p>Would be great to have a confirmation of that</p>",
  "messages": [
    {
      "id": "961574",
      "postDate": "08/07/2020 09:51:33",
      "content": "<p>Looking at the competition rules I see this \n<code>External data, freely &amp; publicly available, is allowed. This includes pre-trained models.</code></p>\n\n<p>I believe this means training a model offline and uploading it for submission is not allowed.</p>\n\n<p>I remember the organizers said in Bengali : \n<code>External data is allowed, and you are encouraged to train your model offline and use your Notebook for inference.</code></p>\n\n<p>Only models trained inside of a notebook are legit ? </p>\n\n<p>Would be great to have a confirmation of that</p>",
      "rawMarkdown": "Looking at the competition rules I see this \n`External data, freely &amp; publicly available, is allowed. This includes pre-trained models.`\n\nI believe this means training a model offline and uploading it for submission is not allowed.\n\nI remember the organizers said in Bengali : \n`External data is allowed, and you are encouraged to train your model offline and use your Notebook for inference.`\n\nOnly models trained inside of a notebook are legit ? \n\nWould be great to have a confirmation of that",
      "votes": null
    },
    {
      "id": "962102",
      "postDate": "08/07/2020 20:04:28",
      "content": "<p>I am new to the data science but I can tell that you have to make submissions through the Kaggle notebooks only. Then only it will be legit</p>",
      "rawMarkdown": "I am new to the data science but I can tell that you have to make submissions through the Kaggle notebooks only. Then only it will be legit",
      "votes": null
    },
    {
      "id": "962462",
      "postDate": "08/08/2020 06:25:39",
      "content": "<p><a href=\"/philculliton\">@philculliton</a> can you please confirm the following 3 notebooks are ALL legit : </p>\n\n<p>Model trained in a notebook:\n - <a href=\"https://www.kaggle.com/akensert/glret-cosface-with-distributed-tf\">https://www.kaggle.com/akensert/glret-cosface-with-distributed-tf</a></p>\n\n<p>Baseline model is publicly available:\n - <a href=\"https://www.kaggle.com/camaskew/baseline-submission\">https://www.kaggle.com/camaskew/baseline-submission</a></p>\n\n<p>Model trained offline, not publicly available. notebook only used to create submission.zip file\n - <a href=\"https://www.kaggle.com/chandanverma/landmark-retrieval-efficientnet\">https://www.kaggle.com/chandanverma/landmark-retrieval-efficientnet</a></p>\n\n<p>Thanks for your help.</p>",
      "rawMarkdown": "philculliton can you please confirm the following 3 notebooks are ALL legit : \n\nModel trained in a notebook:\n - https://www.kaggle.com/akensert/glret-cosface-with-distributed-tf\n\nBaseline model is publicly available:\n - https://www.kaggle.com/camaskew/baseline-submission\n\nModel trained offline, not publicly available. notebook only used to create submission.zip file\n - https://www.kaggle.com/chandanverma/landmark-retrieval-efficientnet\n\nThanks for your help.",
      "votes": null
    },
    {
      "id": "963168",
      "postDate": "08/08/2020 18:11:18",
      "content": "<p>Confirming that all three are OK as long as you state what external data you used.</p>",
      "rawMarkdown": "Confirming that all three are OK as long as you state what external data you used.",
      "votes": null
    },
    {
      "id": "963852",
      "postDate": "08/09/2020 11:07:16",
      "content": "<p>Thanks for your swift confirmation <a href=\"/tobwey\">@tobwey</a> !</p>",
      "rawMarkdown": "Thanks for your swift confirmation @tobwey !",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 962102,
      "author_name": "ajaysinghpanwar",
      "author_url": "",
      "post_date": "08/07/2020 20:04:28",
      "content": "<p>I am new to the data science but I can tell that you have to make submissions through the Kaggle notebooks only. Then only it will be legit</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 962462,
      "author_name": "ogrellier",
      "author_url": "",
      "post_date": "08/08/2020 06:25:39",
      "content": "<p><a href=\"/philculliton\">@philculliton</a> can you please confirm the following 3 notebooks are ALL legit : </p>\n\n<p>Model trained in a notebook:\n - <a href=\"https://www.kaggle.com/akensert/glret-cosface-with-distributed-tf\">https://www.kaggle.com/akensert/glret-cosface-with-distributed-tf</a></p>\n\n<p>Baseline model is publicly available:\n - <a href=\"https://www.kaggle.com/camaskew/baseline-submission\">https://www.kaggle.com/camaskew/baseline-submission</a></p>\n\n<p>Model trained offline, not publicly available. notebook only used to create submission.zip file\n - <a href=\"https://www.kaggle.com/chandanverma/landmark-retrieval-efficientnet\">https://www.kaggle.com/chandanverma/landmark-retrieval-efficientnet</a></p>\n\n<p>Thanks for your help.</p>",
      "votes": null,
      "replies": [
        {
          "id": 963168,
          "author_name": "tobwey",
          "author_url": "",
          "post_date": "08/08/2020 18:11:18",
          "content": "<p>Confirming that all three are OK as long as you state what external data you used.</p>",
          "votes": null,
          "replies": [
            {
              "id": 963852,
              "author_name": "ogrellier",
              "author_url": "",
              "post_date": "08/09/2020 11:07:16",
              "content": "<p>Thanks for your swift confirmation <a href=\"/tobwey\">@tobwey</a> !</p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "961574": "Looking at the competition rules I see this \n`External data, freely &amp; publicly available, is allowed. This includes pre-trained models.`\n\nI believe this means training a model offline and uploading it for submission is not allowed.\n\nI remember the organizers said in Bengali : \n`External data is allowed, and you are encouraged to train your model offline and use your Notebook for inference.`\n\nOnly models trained inside of a notebook are legit ? \n\nWould be great to have a confirmation of that",
    "962102": "I am new to the data science but I can tell that you have to make submissions through the Kaggle notebooks only. Then only it will be legit",
    "962462": "philculliton can you please confirm the following 3 notebooks are ALL legit : \n\nModel trained in a notebook:\n - https://www.kaggle.com/akensert/glret-cosface-with-distributed-tf\n\nBaseline model is publicly available:\n - https://www.kaggle.com/camaskew/baseline-submission\n\nModel trained offline, not publicly available. notebook only used to create submission.zip file\n - https://www.kaggle.com/chandanverma/landmark-retrieval-efficientnet\n\nThanks for your help.",
    "963168": "Confirming that all three are OK as long as you state what external data you used.",
    "963852": "Thanks for your swift confirmation @tobwey !"
  },
  "source": "meta"
}