{
  "id": 149711,
  "title": "Highest score without external data",
  "url": "/competitions/flower-classification-with-tpus/discussion/149711",
  "author_name": "",
  "post_date": "2020-05-09T14:48:32.871449300Z",
  "votes": 3,
  "comment_count": 10,
  "views": 0,
  "content": "<p>Given that Kaggle <a href=\"https://www.kaggle.com/c/flower-classification-with-tpus/discussion/148329#836349\">forbade</a> usage of external data, I suggest to share your highest score achieved without usage of external data in this thread (it seems that public LB is useless now).</p>",
  "messages": [
    {
      "id": "839703",
      "postDate": "05/09/2020 14:48:32",
      "content": "<p>Given that Kaggle <a href=\"https://www.kaggle.com/c/flower-classification-with-tpus/discussion/148329#836349\">forbade</a> usage of external data, I suggest to share your highest score achieved without usage of external data in this thread (it seems that public LB is useless now).</p>",
      "rawMarkdown": "Given that Kaggle [forbade](https://www.kaggle.com/c/flower-classification-with-tpus/discussion/148329#836349) usage of external data, I suggest to share your highest score achieved without usage of external data in this thread (it seems that public LB is useless now).",
      "votes": null
    },
    {
      "id": "839705",
      "postDate": "05/09/2020 14:48:56",
      "content": "<p>Mine is 0.96641</p>",
      "rawMarkdown": "Mine is 0.96641",
      "votes": null
    },
    {
      "id": "839956",
      "postDate": "05/09/2020 17:19:11",
      "content": "<p>Mine is 0.95769 no external data is used.</p>",
      "rawMarkdown": "Mine is 0.95769 no external data is used.",
      "votes": null
    },
    {
      "id": "840762",
      "postDate": "05/10/2020 11:03:24",
      "content": "<p>Mine is 0.96230 (no ensemble)</p>",
      "rawMarkdown": "Mine is 0.96230 (no ensemble)",
      "votes": null
    },
    {
      "id": "841040",
      "postDate": "05/10/2020 14:51:39",
      "content": "<p>Mine is 0.96640, no external data is used.</p>",
      "rawMarkdown": "Mine is 0.96640, no external data is used.",
      "votes": null
    },
    {
      "id": "843581",
      "postDate": "05/12/2020 06:15:22",
      "content": "<p>Mine are\nPrivate 0.96642 Public 0.96739 (w/o external data)\nPrivate 0.98734 Public 0.97759 (w/ external data)</p>\n\n<p>I just selected the highest Public LBs, w/o and w/ external data each.</p>",
      "rawMarkdown": "Mine are\nPrivate 0.96642 Public 0.96739 (w/o external data)\nPrivate 0.98734 Public 0.97759 (w/ external data)\n\nI just selected the highest Public LBs, w/o and w/ external data each.",
      "votes": null
    },
    {
      "id": "844328",
      "postDate": "05/12/2020 15:01:04",
      "content": "<p>Mine is 0.97113 no external data is used.</p>",
      "rawMarkdown": "Mine is 0.97113 no external data is used.",
      "votes": null
    },
    {
      "id": "844492",
      "postDate": "05/12/2020 16:32:46",
      "content": "<p>Private LB 0.96935 with zero external data (from Public LB 0.96851)\nMaybe this info will help the host.</p>",
      "rawMarkdown": "Private LB 0.96935 with zero external data (from Public LB 0.96851)\nMaybe this info will help the host.",
      "votes": null
    },
    {
      "id": "844962",
      "postDate": "05/13/2020 01:04:28",
      "content": "<p><a href=\"/qinhui1999\">@qinhui1999</a>, could you please publish your solution? It seems to be that you are the best without external data.</p>",
      "rawMarkdown": "qinhui1999, could you please publish your solution? It seems to be that you are the best without external data.",
      "votes": null
    },
    {
      "id": "844997",
      "postDate": "05/13/2020 01:56:26",
      "content": "<p><a href=\"/kirillblinov\">@kirillblinov</a> Sure.  The 0.97113 score is a blending kernel.  I will public a kernel which 's private score is 0.97078. <br> Here it is  <a href=\"https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments\">https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments </a></p>",
      "rawMarkdown": "kirillblinov Sure.  The 0.97113 score is a blending kernel.  I will public a kernel which 's private score is 0.97078. <br> Here it is  [https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments ](https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments)",
      "votes": null
    },
    {
      "id": "845055",
      "postDate": "05/13/2020 03:00:44",
      "content": "<p><a href=\"/qinhui1999\">@qinhui1999</a>, thank you. I haven't even tried the random blockout augmentation.</p>",
      "rawMarkdown": "qinhui1999, thank you. I haven't even tried the random blockout augmentation.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 839705,
      "author_name": "atamazian",
      "author_url": "",
      "post_date": "05/09/2020 14:48:56",
      "content": "<p>Mine is 0.96641</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 839956,
      "author_name": "lucca9211",
      "author_url": "",
      "post_date": "05/09/2020 17:19:11",
      "content": "<p>Mine is 0.95769 no external data is used.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 840762,
      "author_name": "ilosvigil",
      "author_url": "",
      "post_date": "05/10/2020 11:03:24",
      "content": "<p>Mine is 0.96230 (no ensemble)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 841040,
      "author_name": "jummatsu",
      "author_url": "",
      "post_date": "05/10/2020 14:51:39",
      "content": "<p>Mine is 0.96640, no external data is used.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 843581,
      "author_name": "kozistr",
      "author_url": "",
      "post_date": "05/12/2020 06:15:22",
      "content": "<p>Mine are\nPrivate 0.96642 Public 0.96739 (w/o external data)\nPrivate 0.98734 Public 0.97759 (w/ external data)</p>\n\n<p>I just selected the highest Public LBs, w/o and w/ external data each.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 844328,
      "author_name": "qinhui1999",
      "author_url": "",
      "post_date": "05/12/2020 15:01:04",
      "content": "<p>Mine is 0.97113 no external data is used.</p>",
      "votes": null,
      "replies": [
        {
          "id": 844962,
          "author_name": "kirillblinov",
          "author_url": "",
          "post_date": "05/13/2020 01:04:28",
          "content": "<p><a href=\"/qinhui1999\">@qinhui1999</a>, could you please publish your solution? It seems to be that you are the best without external data.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 844997,
          "author_name": "qinhui1999",
          "author_url": "",
          "post_date": "05/13/2020 01:56:26",
          "content": "<p><a href=\"/kirillblinov\">@kirillblinov</a> Sure.  The 0.97113 score is a blending kernel.  I will public a kernel which 's private score is 0.97078. <br> Here it is  <a href=\"https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments\">https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments </a></p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 845055,
          "author_name": "kirillblinov",
          "author_url": "",
          "post_date": "05/13/2020 03:00:44",
          "content": "<p><a href=\"/qinhui1999\">@qinhui1999</a>, thank you. I haven't even tried the random blockout augmentation.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 844492,
      "author_name": "coreacasa",
      "author_url": "",
      "post_date": "05/12/2020 16:32:46",
      "content": "<p>Private LB 0.96935 with zero external data (from Public LB 0.96851)\nMaybe this info will help the host.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "839703": "Given that Kaggle [forbade](https://www.kaggle.com/c/flower-classification-with-tpus/discussion/148329#836349) usage of external data, I suggest to share your highest score achieved without usage of external data in this thread (it seems that public LB is useless now).",
    "839705": "Mine is 0.96641",
    "839956": "Mine is 0.95769 no external data is used.",
    "840762": "Mine is 0.96230 (no ensemble)",
    "841040": "Mine is 0.96640, no external data is used.",
    "843581": "Mine are\nPrivate 0.96642 Public 0.96739 (w/o external data)\nPrivate 0.98734 Public 0.97759 (w/ external data)\n\nI just selected the highest Public LBs, w/o and w/ external data each.",
    "844328": "Mine is 0.97113 no external data is used.",
    "844492": "Private LB 0.96935 with zero external data (from Public LB 0.96851)\nMaybe this info will help the host.",
    "844962": "qinhui1999, could you please publish your solution? It seems to be that you are the best without external data.",
    "844997": "kirillblinov Sure.  The 0.97113 score is a blending kernel.  I will public a kernel which 's private score is 0.97078. <br> Here it is  [https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments ](https://www.kaggle.com/qinhui1999/flower-classification-enet-b7-densenet-aug/comments)",
    "845055": "qinhui1999, thank you. I haven't even tried the random blockout augmentation."
  },
  "source": "meta"
}