{
  "id": 108064,
  "title": "Top Performers Size of Training Data and System Configurations",
  "url": "/competitions/aptos2019-blindness-detection/discussion/108064",
  "author_name": "Himanshu Gamit",
  "post_date": "2019-09-08T20:17:08.463000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Congratulations to winners and the participants (same as me) who learned a lot through this competition!!. It is nice to see Private Leaderboard coming up real quick. For me, Model training was done through Kaggle kernels. It would be nice if top performers could shed light on the below questions.</p>\n\n<p><strong>1. What was the size of the training set from the 2015 dataset?</strong>\n*<em>2. What machine configuration you used to train models? Model, Epochs and Training time?</em>*</p>\n\n<p>Thanks </p>",
  "messages": [
    {
      "id": 621669,
      "postDate": "2019-09-08T20:17:08.463Z",
      "content": "<p>Congratulations to winners and the participants (same as me) who learned a lot through this competition!!. It is nice to see Private Leaderboard coming up real quick. For me, Model training was done through Kaggle kernels. It would be nice if top performers could shed light on the below questions.</p>\n\n<p><strong>1. What was the size of the training set from the 2015 dataset?</strong>\n*<em>2. What machine configuration you used to train models? Model, Epochs and Training time?</em>*</p>\n\n<p>Thanks </p>",
      "rawMarkdown": "Congratulations to winners and the participants (same as me) who learned a lot through this competition!!. It is nice to see Private Leaderboard coming up real quick. For me, Model training was done through Kaggle kernels. It would be nice if top performers could shed light on the below questions.\n\n**1. What was the size of the training set from the 2015 dataset?**\n**2. What machine configuration you used to train models? Model, Epochs and Training time?**\n\nThanks \n"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "621669": "Congratulations to winners and the participants (same as me) who learned a lot through this competition!!. It is nice to see Private Leaderboard coming up real quick. For me, Model training was done through Kaggle kernels. It would be nice if top performers could shed light on the below questions.\n\n**1. What was the size of the training set from the 2015 dataset?**\n**2. What machine configuration you used to train models? Model, Epochs and Training time?**\n\nThanks \n"
  }
}