{
  "id": 38547,
  "title": "another benchmark with tensorboard visualization",
  "url": "/competitions/carvana-image-masking-challenge/discussion/38547",
  "author_name": "Jiwei Liu",
  "post_date": "2017-08-25T06:12:01.483000",
  "votes": 8,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi all, here is another benchmark using tensorflow. <br>\n<a href=\"https://github.com/daxiongshu/kaggle-review/tree/master/comps/carvana\">https://github.com/daxiongshu/kaggle-review/tree/master/comps/carvana</a></p>\n\n<p>The lb score is 0.975 with 224x128 resolution. The net is ZFTurbo's unet <a href=\"https://github.com/ZFTurbo/ZF_UNET_224_Pretrained_Model\">https://github.com/ZFTurbo/ZF_UNET_224_Pretrained_Model</a></p>\n\n<p>The main feature is the full online tensorboard visualization including learning curve, the samples of input and predicted mask and the distribution of gradients. Basically you can watch your model's behavior in the browser when it runs. Examples of visualization can be found here:\n<a href=\"https://docs.google.com/presentation/d/1t-bcJEstHGWcd-5LF7IqFFx6T0Kdbt73ZSI05CCryYU/edit?usp=sharing\">https://docs.google.com/presentation/d/1t-bcJEstHGWcd-5LF7IqFFx6T0Kdbt73ZSI05CCryYU/edit?usp=sharing</a></p>\n\n<p>Any feedback is appreciated. Cheers : )</p>",
  "messages": [
    {
      "id": 216292,
      "postDate": "2017-08-25T06:12:01.483Z",
      "content": "<p>Hi all, here is another benchmark using tensorflow. <br>\n<a href=\"https://github.com/daxiongshu/kaggle-review/tree/master/comps/carvana\">https://github.com/daxiongshu/kaggle-review/tree/master/comps/carvana</a></p>\n\n<p>The lb score is 0.975 with 224x128 resolution. The net is ZFTurbo's unet <a href=\"https://github.com/ZFTurbo/ZF_UNET_224_Pretrained_Model\">https://github.com/ZFTurbo/ZF_UNET_224_Pretrained_Model</a></p>\n\n<p>The main feature is the full online tensorboard visualization including learning curve, the samples of input and predicted mask and the distribution of gradients. Basically you can watch your model's behavior in the browser when it runs. Examples of visualization can be found here:\n<a href=\"https://docs.google.com/presentation/d/1t-bcJEstHGWcd-5LF7IqFFx6T0Kdbt73ZSI05CCryYU/edit?usp=sharing\">https://docs.google.com/presentation/d/1t-bcJEstHGWcd-5LF7IqFFx6T0Kdbt73ZSI05CCryYU/edit?usp=sharing</a></p>\n\n<p>Any feedback is appreciated. Cheers : )</p>",
      "rawMarkdown": "Hi all, here is another benchmark using tensorflow.  \nhttps://github.com/daxiongshu/kaggle-review/tree/master/comps/carvana\n\nThe lb score is 0.975 with 224x128 resolution. The net is ZFTurbo's unet https://github.com/ZFTurbo/ZF_UNET_224_Pretrained_Model\n\nThe main feature is the full online tensorboard visualization including learning curve, the samples of input and predicted mask and the distribution of gradients. Basically you can watch your model's behavior in the browser when it runs. Examples of visualization can be found here:\nhttps://docs.google.com/presentation/d/1t-bcJEstHGWcd-5LF7IqFFx6T0Kdbt73ZSI05CCryYU/edit?usp=sharing\n\nAny feedback is appreciated. Cheers : )\n\n\n",
      "votes": 8
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "216292": "Hi all, here is another benchmark using tensorflow.  \nhttps://github.com/daxiongshu/kaggle-review/tree/master/comps/carvana\n\nThe lb score is 0.975 with 224x128 resolution. The net is ZFTurbo's unet https://github.com/ZFTurbo/ZF_UNET_224_Pretrained_Model\n\nThe main feature is the full online tensorboard visualization including learning curve, the samples of input and predicted mask and the distribution of gradients. Basically you can watch your model's behavior in the browser when it runs. Examples of visualization can be found here:\nhttps://docs.google.com/presentation/d/1t-bcJEstHGWcd-5LF7IqFFx6T0Kdbt73ZSI05CCryYU/edit?usp=sharing\n\nAny feedback is appreciated. Cheers : )\n\n\n"
  }
}