{
  "id": 157819,
  "title": "Starter Tensorflow Kernel",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/157819",
  "author_name": "Durgesh",
  "post_date": "2020-06-12T06:55:27.691000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<h2>Melanoma Detection</h2>\n\n<p>KERNEL: <a href=\"https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu\">https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu</a></p>\n\n<p>Bare Bone Modularized code written in tensorflow 2.x for melanoma detection It could be good starting point to implement your strategy on top of it. In this kernel I tried to build a pipeline that is easy to maintain and provide greater extend to modification</p>\n\n<p><strong>Note:</strong> This kernel works on any hardware out of the box without any extra configuration</p>\n\n<h3>Things that are Implemented</h3>\n\n<p>[TFrec-loader] --&gt; [tf.Data.Dataset()] --&gt; [PreProcess functions] --&gt; [tf sequential model with pretrained weights] --&gt; [training instructions]</p>\n\n<h3>Things that can be strategically implemented</h3>\n\n<ul>\n<li>Cross Validation[Folds]</li>\n<li>Augmentation</li>\n<li>Model Tuning</li>\n<li>etc,</li>\n</ul>",
  "messages": [
    {
      "id": 882833,
      "postDate": "2020-06-12T06:55:27.690Z",
      "content": "<h2>Melanoma Detection</h2>\n\n<p>KERNEL: <a href=\"https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu\">https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu</a></p>\n\n<p>Bare Bone Modularized code written in tensorflow 2.x for melanoma detection It could be good starting point to implement your strategy on top of it. In this kernel I tried to build a pipeline that is easy to maintain and provide greater extend to modification</p>\n\n<p><strong>Note:</strong> This kernel works on any hardware out of the box without any extra configuration</p>\n\n<h3>Things that are Implemented</h3>\n\n<p>[TFrec-loader] --&gt; [tf.Data.Dataset()] --&gt; [PreProcess functions] --&gt; [tf sequential model with pretrained weights] --&gt; [training instructions]</p>\n\n<h3>Things that can be strategically implemented</h3>\n\n<ul>\n<li>Cross Validation[Folds]</li>\n<li>Augmentation</li>\n<li>Model Tuning</li>\n<li>etc,</li>\n</ul>",
      "rawMarkdown": "\n## Melanoma Detection\n\nKERNEL: [https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu](https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu)\n\nBare Bone Modularized code written in tensorflow 2.x for melanoma detection It could be good starting point to implement your strategy on top of it. In this kernel I tried to build a pipeline that is easy to maintain and provide greater extend to modification\n\n**Note:** This kernel works on any hardware out of the box without any extra configuration\n\n### Things that are Implemented\n\n[TFrec-loader] --&gt; [tf.Data.Dataset()] --&gt; [PreProcess functions] --&gt; [tf sequential model with pretrained weights] --&gt; [training instructions]\n\n\n### Things that can be strategically implemented\n\n- Cross Validation[Folds]\n- Augmentation\n- Model Tuning\n- etc,\n\n"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "882833": "\n## Melanoma Detection\n\nKERNEL: [https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu](https://www.kaggle.com/orionpax00/starter-kernel-melanoma-tf2-0-tpu)\n\nBare Bone Modularized code written in tensorflow 2.x for melanoma detection It could be good starting point to implement your strategy on top of it. In this kernel I tried to build a pipeline that is easy to maintain and provide greater extend to modification\n\n**Note:** This kernel works on any hardware out of the box without any extra configuration\n\n### Things that are Implemented\n\n[TFrec-loader] --&gt; [tf.Data.Dataset()] --&gt; [PreProcess functions] --&gt; [tf sequential model with pretrained weights] --&gt; [training instructions]\n\n\n### Things that can be strategically implemented\n\n- Cross Validation[Folds]\n- Augmentation\n- Model Tuning\n- etc,\n\n"
  }
}