{
  "id": 55559,
  "title": "can kernel handles image data?",
  "url": "/competitions/avito-demand-prediction/discussion/55559",
  "author_name": "",
  "post_date": "2018-04-28T15:07:14.657808Z",
  "votes": 1,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Image data is so big, can handle image datas in kernel?</p>",
  "messages": [
    {
      "id": "320396",
      "postDate": "04/28/2018 15:07:14",
      "content": "<p>Image data is so big, can handle image datas in kernel?</p>",
      "rawMarkdown": "Image data is so big, can handle image datas in kernel?",
      "votes": null
    },
    {
      "id": "320533",
      "postDate": "04/29/2018 00:44:35",
      "content": "<p>It depends on your exact model, but I've built a kernel on Kaggle that's worked with more image data than this.</p>",
      "rawMarkdown": "It depends on your exact model, but I've built a kernel on Kaggle that's worked with more image data than this.",
      "votes": null
    },
    {
      "id": "320540",
      "postDate": "04/29/2018 02:09:13",
      "content": "<p>wow!Great,I will try some simple model</p>",
      "rawMarkdown": "wow!Great,I will try some simple model",
      "votes": null
    },
    {
      "id": "320788",
      "postDate": "04/29/2018 20:58:03",
      "content": "<p>Idk how you're going about training but you could use keras's fit_generator feature. A link that goes over what i'm talking about \n<a href=\"https://stanford.edu/~shervine/blog/keras-how-to-generate-data-on-the-fly.html\">https://stanford.edu/~shervine/blog/keras-how-to-generate-data-on-the-fly.html</a></p>",
      "rawMarkdown": "Idk how you're going about training but you could use keras's fit_generator feature. A link that goes over what i'm talking about \nhttps://stanford.edu/~shervine/blog/keras-how-to-generate-data-on-the-fly.html",
      "votes": null
    },
    {
      "id": "320843",
      "postDate": "04/30/2018 01:56:51",
      "content": "<p>Several of the existing kernels involve image data.</p>",
      "rawMarkdown": "Several of the existing kernels involve image data.",
      "votes": null
    },
    {
      "id": "321039",
      "postDate": "04/30/2018 13:00:24",
      "content": "<p>Thank you for your post.I think now the only question to handle all images in kernels is  time limitation,I will try to handle image datas in several kernels.</p>",
      "rawMarkdown": "Thank you for your post.I think now the only question to handle all images in kernels is  time limitation,I will try to handle image datas in several kernels.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 320533,
      "author_name": "matthewa313",
      "author_url": "",
      "post_date": "04/29/2018 00:44:35",
      "content": "<p>It depends on your exact model, but I've built a kernel on Kaggle that's worked with more image data than this.</p>",
      "votes": null,
      "replies": [
        {
          "id": 320540,
          "author_name": "liuhdsgoal",
          "author_url": "",
          "post_date": "04/29/2018 02:09:13",
          "content": "<p>wow!Great,I will try some simple model</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 320788,
      "author_name": "wobeert",
      "author_url": "",
      "post_date": "04/29/2018 20:58:03",
      "content": "<p>Idk how you're going about training but you could use keras's fit_generator feature. A link that goes over what i'm talking about \n<a href=\"https://stanford.edu/~shervine/blog/keras-how-to-generate-data-on-the-fly.html\">https://stanford.edu/~shervine/blog/keras-how-to-generate-data-on-the-fly.html</a></p>",
      "votes": null,
      "replies": [
        {
          "id": 321039,
          "author_name": "liuhdsgoal",
          "author_url": "",
          "post_date": "04/30/2018 13:00:24",
          "content": "<p>Thank you for your post.I think now the only question to handle all images in kernels is  time limitation,I will try to handle image datas in several kernels.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 320843,
      "author_name": "jpmiller",
      "author_url": "",
      "post_date": "04/30/2018 01:56:51",
      "content": "<p>Several of the existing kernels involve image data.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "320396": "Image data is so big, can handle image datas in kernel?",
    "320533": "It depends on your exact model, but I've built a kernel on Kaggle that's worked with more image data than this.",
    "320540": "wow!Great,I will try some simple model",
    "320788": "Idk how you're going about training but you could use keras's fit_generator feature. A link that goes over what i'm talking about \nhttps://stanford.edu/~shervine/blog/keras-how-to-generate-data-on-the-fly.html",
    "320843": "Several of the existing kernels involve image data.",
    "321039": "Thank you for your post.I think now the only question to handle all images in kernels is  time limitation,I will try to handle image datas in several kernels."
  },
  "source": "meta"
}