{
  "id": 57059,
  "title": "dataset in train_color is more then 90 GB volume ",
  "url": "/competitions/cvpr-2018-autonomous-driving/discussion/57059",
  "author_name": "",
  "post_date": "2018-05-18T15:10:46.236548900Z",
  "votes": null,
  "comment_count": 4,
  "views": 0,
  "content": "<p>this their any way to work with the train_color data without downloading train_color.zip </p>",
  "messages": [
    {
      "id": "330320",
      "postDate": "05/18/2018 15:10:46",
      "content": "<p>this their any way to work with the train_color data without downloading train_color.zip </p>",
      "rawMarkdown": "this their any way to work with the train_color data without downloading train_color.zip",
      "votes": null
    },
    {
      "id": "331314",
      "postDate": "05/21/2018 00:31:55",
      "content": "<p>I don't think there is a point in participating in the competition without using the training data. After all, how would you train your model? </p>",
      "rawMarkdown": "I don't think there is a point in participating in the competition without using the training data. After all, how would you train your model?",
      "votes": null
    },
    {
      "id": "331364",
      "postDate": "05/21/2018 05:14:15",
      "content": "<p>hi, my problem is not about using  training data. My internet package is limited to 4 GB per day and won't be able to download complete train_color. Can anyone be able to share atleast 10,000 train_color image dataset , so that i can download and work with that.</p>",
      "rawMarkdown": "hi, my problem is not about using  training data. My internet package is limited to 4 GB per day and won't be able to download complete train_color. Can anyone be able to share atleast 10,000 train_color image dataset , so that i can download and work with that.",
      "votes": null
    },
    {
      "id": "331373",
      "postDate": "05/21/2018 05:29:03",
      "content": "<p>anyway, I am going to change the boardband connection for permanent solution. </p>",
      "rawMarkdown": "anyway, I am going to change the boardband connection for permanent solution.",
      "votes": null
    },
    {
      "id": "331644",
      "postDate": "05/21/2018 16:18:37",
      "content": "<p>The images have a resolution of  3384x2710, so those 4GB won't get you very far, based on the average size of 2.5 MB you will only get 1600 images</p>",
      "rawMarkdown": "The images have a resolution of  3384x2710, so those 4GB won't get you very far, based on the average size of 2.5 MB you will only get 1600 images",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 331314,
      "author_name": "michaelheinzer",
      "author_url": "",
      "post_date": "05/21/2018 00:31:55",
      "content": "<p>I don't think there is a point in participating in the competition without using the training data. After all, how would you train your model? </p>",
      "votes": null,
      "replies": [
        {
          "id": 331364,
          "author_name": "",
          "author_url": "",
          "post_date": "05/21/2018 05:14:15",
          "content": "<p>hi, my problem is not about using  training data. My internet package is limited to 4 GB per day and won't be able to download complete train_color. Can anyone be able to share atleast 10,000 train_color image dataset , so that i can download and work with that.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 331373,
          "author_name": "",
          "author_url": "",
          "post_date": "05/21/2018 05:29:03",
          "content": "<p>anyway, I am going to change the boardband connection for permanent solution. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 331644,
          "author_name": "michaelheinzer",
          "author_url": "",
          "post_date": "05/21/2018 16:18:37",
          "content": "<p>The images have a resolution of  3384x2710, so those 4GB won't get you very far, based on the average size of 2.5 MB you will only get 1600 images</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "330320": "this their any way to work with the train_color data without downloading train_color.zip",
    "331314": "I don't think there is a point in participating in the competition without using the training data. After all, how would you train your model?",
    "331364": "hi, my problem is not about using  training data. My internet package is limited to 4 GB per day and won't be able to download complete train_color. Can anyone be able to share atleast 10,000 train_color image dataset , so that i can download and work with that.",
    "331373": "anyway, I am going to change the boardband connection for permanent solution.",
    "331644": "The images have a resolution of  3384x2710, so those 4GB won't get you very far, based on the average size of 2.5 MB you will only get 1600 images"
  },
  "source": "meta"
}