{
  "id": 174012,
  "title": "Results of models other than efficientnet.",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/174012",
  "author_name": "",
  "post_date": "2020-08-11T19:55:47.808506700Z",
  "votes": 2,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Feel free to share your thoughts on other models like Xception, VGG, ResNet, Inception, MobileNet etc.</p>",
  "messages": [
    {
      "id": "966965",
      "postDate": "08/11/2020 19:55:47",
      "content": "<p>Feel free to share your thoughts on other models like Xception, VGG, ResNet, Inception, MobileNet etc.</p>",
      "rawMarkdown": "Feel free to share your thoughts on other models like Xception, VGG, ResNet, Inception, MobileNet etc.",
      "votes": null
    },
    {
      "id": "967226",
      "postDate": "08/12/2020 05:33:10",
      "content": "<p>My train 5 folds (change parameters and configs + <a href=\"https://www.kaggle.com/truonghoang/multi-size-eff-lb-0-912/comments\" target=\"_blank\">https://www.kaggle.com/truonghoang/multi-size-eff-lb-0-912/comments</a>) : VGG 0.87, Inception 0.92. some other backbones get 0.92 too</p>",
      "rawMarkdown": "My train 5 folds (change parameters and configs + https://www.kaggle.com/truonghoang/multi-size-eff-lb-0-912/comments) : VGG 0.87, Inception 0.92. some other backbones get 0.92 too",
      "votes": null
    },
    {
      "id": "967227",
      "postDate": "08/12/2020 05:33:21",
      "content": "<p>Nice One </p>",
      "rawMarkdown": "Nice One",
      "votes": null
    },
    {
      "id": "968480",
      "postDate": "08/13/2020 04:01:51",
      "content": "<p>I have tried all the models. I observed EfficientNet &gt; DenseNet &gt; Xception &gt; ResNet &gt; Inception &gt; VGG</p>",
      "rawMarkdown": "I have tried all the models. I observed EfficientNet > DenseNet > Xception > ResNet > Inception > VGG",
      "votes": null
    },
    {
      "id": "969465",
      "postDate": "08/13/2020 17:48:25",
      "content": "<p>Thanks for sharing. Can you also post CV, LB scores of each of these or a few of these?</p>",
      "rawMarkdown": "Thanks for sharing. Can you also post CV, LB scores of each of these or a few of these?",
      "votes": null
    },
    {
      "id": "969505",
      "postDate": "08/13/2020 18:22:26",
      "content": "<p>For me, </p>\n<pre><code>DenseNet &gt; GhostNet &gt; SeResNeXT50 &gt; MobileNet &gt; TResNet\n</code></pre>\n<p><strong>EfficientNet</strong> is pretty efficient I guess. -)</p>",
      "rawMarkdown": "For me, \n\n```\nDenseNet > GhostNet > SeResNeXT50 > MobileNet > TResNet\n```\n**EfficientNet** is pretty efficient I guess. -)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 967226,
      "author_name": "truonghoang",
      "author_url": "",
      "post_date": "08/12/2020 05:33:10",
      "content": "<p>My train 5 folds (change parameters and configs + <a href=\"https://www.kaggle.com/truonghoang/multi-size-eff-lb-0-912/comments\" target=\"_blank\">https://www.kaggle.com/truonghoang/multi-size-eff-lb-0-912/comments</a>) : VGG 0.87, Inception 0.92. some other backbones get 0.92 too</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 968480,
      "author_name": "cdeotte",
      "author_url": "",
      "post_date": "08/13/2020 04:01:51",
      "content": "<p>I have tried all the models. I observed EfficientNet &gt; DenseNet &gt; Xception &gt; ResNet &gt; Inception &gt; VGG</p>",
      "votes": null,
      "replies": [
        {
          "id": 969465,
          "author_name": "vikrant06",
          "author_url": "",
          "post_date": "08/13/2020 17:48:25",
          "content": "<p>Thanks for sharing. Can you also post CV, LB scores of each of these or a few of these?</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 969505,
      "author_name": "ipythonx",
      "author_url": "",
      "post_date": "08/13/2020 18:22:26",
      "content": "<p>For me, </p>\n<pre><code>DenseNet &gt; GhostNet &gt; SeResNeXT50 &gt; MobileNet &gt; TResNet\n</code></pre>\n<p><strong>EfficientNet</strong> is pretty efficient I guess. -)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 967227,
      "author_name": "sudarshanpatil",
      "author_url": "",
      "post_date": "08/12/2020 05:33:21",
      "content": "<p>Nice One </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "966965": "Feel free to share your thoughts on other models like Xception, VGG, ResNet, Inception, MobileNet etc.",
    "967226": "My train 5 folds (change parameters and configs + https://www.kaggle.com/truonghoang/multi-size-eff-lb-0-912/comments) : VGG 0.87, Inception 0.92. some other backbones get 0.92 too",
    "967227": "Nice One",
    "968480": "I have tried all the models. I observed EfficientNet > DenseNet > Xception > ResNet > Inception > VGG",
    "969465": "Thanks for sharing. Can you also post CV, LB scores of each of these or a few of these?",
    "969505": "For me, \n\n```\nDenseNet > GhostNet > SeResNeXT50 > MobileNet > TResNet\n```\n**EfficientNet** is pretty efficient I guess. -)"
  },
  "source": "meta"
}