{
  "id": 161965,
  "title": "Gathering the data and training your own model",
  "url": "/competitions/open-images-object-detection-rvc-2020/discussion/161965",
  "author_name": "",
  "post_date": "2020-06-26T20:44:12.011525500Z",
  "votes": 5,
  "comment_count": 1,
  "views": 0,
  "content": "<p>To get on the leaderboard I used the a pretrained object detection model from tf hub. </p>\n\n<p>I would like to move on to actually training my own model.. where should I store the data?</p>\n\n<p>Can I download the data to Kaggle? Or should I download the data to google drive and use google colab and use one of their free GPU/TPU's? </p>\n\n<p>Also I see google cloud offers a $300 credit for new users. Will that be enough to store the train and val data, and train a model?</p>\n\n<p>Any other options I should look into?</p>\n\n<p>@jilian , @golubev , @drn01z3 , @vecxoz , @a2015003713 , @fengepsilon \nhaha sorry for the spam but the discussions in this competition seem a bit dry, you are all doing really great in this competition, and I (and I'm sure many others) would really appreciate some guidance on getting started.</p>\n\n<p>Anyone? Please, I really want to learn more about object detection. Thanks!</p>",
  "messages": [
    {
      "id": "903452",
      "postDate": "06/26/2020 20:44:12",
      "content": "<p>To get on the leaderboard I used the a pretrained object detection model from tf hub. </p>\n\n<p>I would like to move on to actually training my own model.. where should I store the data?</p>\n\n<p>Can I download the data to Kaggle? Or should I download the data to google drive and use google colab and use one of their free GPU/TPU's? </p>\n\n<p>Also I see google cloud offers a $300 credit for new users. Will that be enough to store the train and val data, and train a model?</p>\n\n<p>Any other options I should look into?</p>\n\n<p>@jilian , @golubev , @drn01z3 , @vecxoz , @a2015003713 , @fengepsilon \nhaha sorry for the spam but the discussions in this competition seem a bit dry, you are all doing really great in this competition, and I (and I'm sure many others) would really appreciate some guidance on getting started.</p>\n\n<p>Anyone? Please, I really want to learn more about object detection. Thanks!</p>",
      "rawMarkdown": "To get on the leaderboard I used the a pretrained object detection model from tf hub. \n\nI would like to move on to actually training my own model.. where should I store the data?\n\nCan I download the data to Kaggle? Or should I download the data to google drive and use google colab and use one of their free GPU/TPU's? \n\nAlso I see google cloud offers a $300 credit for new users. Will that be enough to store the train and val data, and train a model?\n\nAny other options I should look into?\n\n@jilian , @golubev , @drn01z3 , @vecxoz , @a2015003713 , @fengepsilon \nhaha sorry for the spam but the discussions in this competition seem a bit dry, you are all doing really great in this competition, and I (and I'm sure many others) would really appreciate some guidance on getting started.\n\nAnyone? Please, I really want to learn more about object detection. Thanks!",
      "votes": null
    },
    {
      "id": "965020",
      "postDate": "08/10/2020 10:30:17",
      "content": "<p>This competition is somewhat difficult because you'll require quite a lot computational resources if you want to train your own model. The best option for you (unless you don't have personal GPU cluster) is to train your model on Google Colab Pro, and when use inference notebook to submit your results.</p>",
      "rawMarkdown": "This competition is somewhat difficult because you'll require quite a lot computational resources if you want to train your own model. The best option for you (unless you don't have personal GPU cluster) is to train your model on Google Colab Pro, and when use inference notebook to submit your results.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 965020,
      "author_name": "atamazian",
      "author_url": "",
      "post_date": "08/10/2020 10:30:17",
      "content": "<p>This competition is somewhat difficult because you'll require quite a lot computational resources if you want to train your own model. The best option for you (unless you don't have personal GPU cluster) is to train your model on Google Colab Pro, and when use inference notebook to submit your results.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "903452": "To get on the leaderboard I used the a pretrained object detection model from tf hub. \n\nI would like to move on to actually training my own model.. where should I store the data?\n\nCan I download the data to Kaggle? Or should I download the data to google drive and use google colab and use one of their free GPU/TPU's? \n\nAlso I see google cloud offers a $300 credit for new users. Will that be enough to store the train and val data, and train a model?\n\nAny other options I should look into?\n\n@jilian , @golubev , @drn01z3 , @vecxoz , @a2015003713 , @fengepsilon \nhaha sorry for the spam but the discussions in this competition seem a bit dry, you are all doing really great in this competition, and I (and I'm sure many others) would really appreciate some guidance on getting started.\n\nAnyone? Please, I really want to learn more about object detection. Thanks!",
    "965020": "This competition is somewhat difficult because you'll require quite a lot computational resources if you want to train your own model. The best option for you (unless you don't have personal GPU cluster) is to train your model on Google Colab Pro, and when use inference notebook to submit your results."
  },
  "source": "meta"
}