{
  "id": 395818,
  "title": "Solution using PyTorch",
  "url": "/competitions/digit-recognizer/discussion/395818",
  "author_name": "Josemaria",
  "post_date": "2023-03-19T02:17:00.989000",
  "votes": 2,
  "comment_count": 2,
  "views": null,
  "content": "<p>Just wanted to leave this here for anyone that is trying to use PyTorch in this competition. I got a 97% accuracy with a CNN, using RMSprop and Cross Entropy. </p>\n<p>I tried to document the code as much as I could, I hope someone finds it useful.</p>\n<p>Here is the link to the <a href=\"https://github.com/jjmmchema/MNISTwPyTorch\" target=\"_blank\">repository</a></p>",
  "messages": [
    {
      "id": 2187727,
      "postDate": "2023-03-19T02:17:00.990Z",
      "content": "<p>Just wanted to leave this here for anyone that is trying to use PyTorch in this competition. I got a 97% accuracy with a CNN, using RMSprop and Cross Entropy. </p>\n<p>I tried to document the code as much as I could, I hope someone finds it useful.</p>\n<p>Here is the link to the <a href=\"https://github.com/jjmmchema/MNISTwPyTorch\" target=\"_blank\">repository</a></p>",
      "rawMarkdown": "Just wanted to leave this here for anyone that is trying to use PyTorch in this competition. I got a 97% accuracy with a CNN, using RMSprop and Cross Entropy. \n\nI tried to document the code as much as I could, I hope someone finds it useful.\n\nHere is the link to the [repository](https://github.com/jjmmchema/MNISTwPyTorch)\n",
      "votes": 2
    },
    {
      "id": 2224452,
      "postDate": "2023-04-17T10:47:53.437Z",
      "content": "<p>Accuracy score?</p>",
      "rawMarkdown": "Accuracy score?\n",
      "replies": [
        {
          "id": 2226522,
          "postDate": "2023-04-19T02:37:14.790Z",
          "content": "<blockquote>\n  <p>Accuracy score?</p>\n</blockquote>\n<p>Got 0.97282</p>",
          "rawMarkdown": "> Accuracy score?\n\nGot 0.97282"
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 2224452,
      "author_name": "Tobias",
      "author_url": "",
      "post_date": "2023-04-17T10:47:53.437000",
      "content": "<p>Accuracy score?</p>",
      "votes": 0,
      "replies": [
        {
          "id": 2226522,
          "author_name": "Josemaria",
          "author_url": "",
          "post_date": "2023-04-19T02:37:14.790000",
          "content": "<blockquote>\n  <p>Accuracy score?</p>\n</blockquote>\n<p>Got 0.97282</p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2187727": "Just wanted to leave this here for anyone that is trying to use PyTorch in this competition. I got a 97% accuracy with a CNN, using RMSprop and Cross Entropy. \n\nI tried to document the code as much as I could, I hope someone finds it useful.\n\nHere is the link to the [repository](https://github.com/jjmmchema/MNISTwPyTorch)\n",
    "2224452": "Accuracy score?\n"
  }
}