{
  "id": 32973,
  "title": "TrainSmall2.7z is enough",
  "url": "/competitions/noaa-fisheries-steller-sea-lion-population-count/discussion/32973",
  "author_name": "",
  "post_date": "2017-05-13T21:24:31.695319800Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Is trainsmall2.7z enough for predicting the output because I dont have that much space to download images of 95 GB</p>",
  "messages": [
    {
      "id": "182440",
      "postDate": "05/13/2017 21:24:31",
      "content": "<p>Is trainsmall2.7z enough for predicting the output because I dont have that much space to download images of 95 GB</p>",
      "rawMarkdown": "Is trainsmall2.7z enough for predicting the output because I dont have that much space to download images of 95 GB",
      "votes": null
    },
    {
      "id": "182477",
      "postDate": "05/14/2017 02:34:26",
      "content": "<p>Hi @NivedithaVittal, </p>\n\n<p>I think since according to the <a href=\"https://www.kaggle.com/c/noaa-fisheries-steller-sea-lion-population-count/data\">data page</a>, TrainSmall2.7z only \"contains the regular and dotted versions of training images 41 to 50\", you will only have 10 images to train on, out of the 891 images available to everyone else you'll be at a significant disadvantage. But I was in your situation and I got a 4 TB external hard drive (you really only need 1 TB), and I can store all the data on it and process virtually the same way! Another option is investing in <a href=\"https://aws.amazon.com/\">amazon instances</a> where you pay for your storage and operation costs. </p>",
      "rawMarkdown": "Hi @NivedithaVittal, \n\nI think since according to the [data page][1], TrainSmall2.7z only \"contains the regular and dotted versions of training images 41 to 50\", you will only have 10 images to train on, out of the 891 images available to everyone else you'll be at a significant disadvantage. But I was in your situation and I got a 4 TB external hard drive (you really only need 1 TB), and I can store all the data on it and process virtually the same way! Another option is investing in [amazon instances][2] where you pay for your storage and operation costs. \n\n\n  [1]: https://www.kaggle.com/c/noaa-fisheries-steller-sea-lion-population-count/data\n  [2]: https://aws.amazon.com/",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 182477,
      "author_name": "livingprogram",
      "author_url": "",
      "post_date": "05/14/2017 02:34:26",
      "content": "<p>Hi @NivedithaVittal, </p>\n\n<p>I think since according to the <a href=\"https://www.kaggle.com/c/noaa-fisheries-steller-sea-lion-population-count/data\">data page</a>, TrainSmall2.7z only \"contains the regular and dotted versions of training images 41 to 50\", you will only have 10 images to train on, out of the 891 images available to everyone else you'll be at a significant disadvantage. But I was in your situation and I got a 4 TB external hard drive (you really only need 1 TB), and I can store all the data on it and process virtually the same way! Another option is investing in <a href=\"https://aws.amazon.com/\">amazon instances</a> where you pay for your storage and operation costs. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "182440": "Is trainsmall2.7z enough for predicting the output because I dont have that much space to download images of 95 GB",
    "182477": "Hi @NivedithaVittal, \n\nI think since according to the [data page][1], TrainSmall2.7z only \"contains the regular and dotted versions of training images 41 to 50\", you will only have 10 images to train on, out of the 891 images available to everyone else you'll be at a significant disadvantage. But I was in your situation and I got a 4 TB external hard drive (you really only need 1 TB), and I can store all the data on it and process virtually the same way! Another option is investing in [amazon instances][2] where you pay for your storage and operation costs. \n\n\n  [1]: https://www.kaggle.com/c/noaa-fisheries-steller-sea-lion-population-count/data\n  [2]: https://aws.amazon.com/"
  },
  "source": "meta"
}