{
  "id": 90129,
  "title": "Accessing the data",
  "url": "/competitions/iwildcam-2019-fgvc6/discussion/90129",
  "author_name": "",
  "post_date": "2019-04-20T18:48:42.759711600Z",
  "votes": 3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>New to kaggle contests any recommendations for accessing the data ? downloading it locally is 43 GB .. is there any way to access images for train with an http request ? or should I use some cloud service for storing the data ?</p>",
  "messages": [
    {
      "id": "520345",
      "postDate": "04/20/2019 18:48:42",
      "content": "<p>New to kaggle contests any recommendations for accessing the data ? downloading it locally is 43 GB .. is there any way to access images for train with an http request ? or should I use some cloud service for storing the data ?</p>",
      "rawMarkdown": "New to kaggle contests any recommendations for accessing the data ? downloading it locally is 43 GB .. is there any way to access images for train with an http request ? or should I use some cloud service for storing the data ?",
      "votes": null
    },
    {
      "id": "521445",
      "postDate": "04/22/2019 23:43:28",
      "content": "<p>Kaggle itself can be used. Try making a kernel for this competition. The train data is provided, as well a Jupyter notebook. Basically, this means that you can have all your data and development on Kaggle. </p>\n\n<p>Included\nDevelopment Envirnoment\nData\nCPU/GPU</p>",
      "rawMarkdown": "Kaggle itself can be used. Try making a kernel for this competition. The train data is provided, as well a Jupyter notebook. Basically, this means that you can have all your data and development on Kaggle. \n\nIncluded\nDevelopment Envirnoment\nData\nCPU/GPU",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 521445,
      "author_name": "milesb",
      "author_url": "",
      "post_date": "04/22/2019 23:43:28",
      "content": "<p>Kaggle itself can be used. Try making a kernel for this competition. The train data is provided, as well a Jupyter notebook. Basically, this means that you can have all your data and development on Kaggle. </p>\n\n<p>Included\nDevelopment Envirnoment\nData\nCPU/GPU</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "520345": "New to kaggle contests any recommendations for accessing the data ? downloading it locally is 43 GB .. is there any way to access images for train with an http request ? or should I use some cloud service for storing the data ?",
    "521445": "Kaggle itself can be used. Try making a kernel for this competition. The train data is provided, as well a Jupyter notebook. Basically, this means that you can have all your data and development on Kaggle. \n\nIncluded\nDevelopment Envirnoment\nData\nCPU/GPU"
  },
  "source": "meta"
}