{
  "id": 215377,
  "title": "Bottleneck Transformers for Visual Recognition",
  "url": "/competitions/hpa-single-cell-image-classification/discussion/215377",
  "author_name": "Mobassir",
  "post_date": "2021-01-29T15:20:01.922000",
  "votes": 8,
  "comment_count": 4,
  "views": 0,
  "content": "<p><img src=\"https://media-exp1.licdn.com/dms/image/C4E22AQGgQ--ueNtBSg/feedshare-shrink_1280/0/1611887346797?e=1614816000&amp;v=beta&amp;t=L-D-uIphcEW35H5iB5tauxmQYzh2H8cU8w8PBdugk6c\" alt=\"\"></p>\n<h1>paper : <a href=\"https://arxiv.org/pdf/2101.11605.pdf\" target=\"_blank\">https://arxiv.org/pdf/2101.11605.pdf</a></h1>\n<h1>code 1 : <a href=\"https://github.com/leaderj1001/BottleneckTransformers\" target=\"_blank\">https://github.com/leaderj1001/BottleneckTransformers</a></h1>\n<h1>code 2 : <a href=\"https://github.com/lucidrains/bottleneck-transformer-pytorch\" target=\"_blank\">https://github.com/lucidrains/bottleneck-transformer-pytorch</a></h1>",
  "messages": [
    {
      "id": 1176317,
      "postDate": "2021-01-29T15:20:01.923Z",
      "content": "<p><img src=\"https://media-exp1.licdn.com/dms/image/C4E22AQGgQ--ueNtBSg/feedshare-shrink_1280/0/1611887346797?e=1614816000&amp;v=beta&amp;t=L-D-uIphcEW35H5iB5tauxmQYzh2H8cU8w8PBdugk6c\" alt=\"\"></p>\n<h1>paper : <a href=\"https://arxiv.org/pdf/2101.11605.pdf\" target=\"_blank\">https://arxiv.org/pdf/2101.11605.pdf</a></h1>\n<h1>code 1 : <a href=\"https://github.com/leaderj1001/BottleneckTransformers\" target=\"_blank\">https://github.com/leaderj1001/BottleneckTransformers</a></h1>\n<h1>code 2 : <a href=\"https://github.com/lucidrains/bottleneck-transformer-pytorch\" target=\"_blank\">https://github.com/lucidrains/bottleneck-transformer-pytorch</a></h1>",
      "rawMarkdown": "![](https://media-exp1.licdn.com/dms/image/C4E22AQGgQ--ueNtBSg/feedshare-shrink_1280/0/1611887346797?e=1614816000&v=beta&t=L-D-uIphcEW35H5iB5tauxmQYzh2H8cU8w8PBdugk6c)\n\n# paper : https://arxiv.org/pdf/2101.11605.pdf\n\n# code 1 : https://github.com/leaderj1001/BottleneckTransformers\n\n# code 2 : https://github.com/lucidrains/bottleneck-transformer-pytorch",
      "votes": 8
    },
    {
      "id": 1178367,
      "postDate": "2021-01-30T20:22:25.167Z",
      "content": "<p>Google research and the code was made in PyTorch, interesting. </p>",
      "rawMarkdown": "Google research and the code was made in PyTorch, interesting. ",
      "votes": 1,
      "replies": [
        {
          "id": 1178531,
          "postDate": "2021-01-31T00:33:42.460Z",
          "content": "<p>The <a href=\"https://gist.github.com/aravindsrinivas/56359b79f0ce4449bcb04ab4b56a57a2\" target=\"_blank\">original implementation</a>, linked from the paper is implemented in tensorflow.</p>",
          "rawMarkdown": "The [original implementation](https://gist.github.com/aravindsrinivas/56359b79f0ce4449bcb04ab4b56a57a2), linked from the paper is implemented in tensorflow.",
          "votes": 2
        },
        {
          "id": 1178566,
          "postDate": "2021-01-31T01:33:28.077Z",
          "content": "<p>Thanks for share</p>",
          "rawMarkdown": "Thanks for share"
        }
      ]
    },
    {
      "id": 1177674,
      "postDate": "2021-01-30T12:58:26.163Z",
      "content": "<p>Superb!!! It's really a considerable job and Keep it up. </p>",
      "rawMarkdown": "Superb!!! It's really a considerable job and Keep it up. "
    }
  ],
  "comments": [
    {
      "id": 1178367,
      "author_name": "Hiram Coria 🧬",
      "author_url": "",
      "post_date": "2021-01-30T20:22:25.167000",
      "content": "<p>Google research and the code was made in PyTorch, interesting. </p>",
      "votes": 1,
      "replies": [
        {
          "id": 1178531,
          "author_name": "Dmytro Poplavskiy",
          "author_url": "",
          "post_date": "2021-01-31T00:33:42.460000",
          "content": "<p>The <a href=\"https://gist.github.com/aravindsrinivas/56359b79f0ce4449bcb04ab4b56a57a2\" target=\"_blank\">original implementation</a>, linked from the paper is implemented in tensorflow.</p>",
          "votes": 2,
          "replies": []
        },
        {
          "id": 1178566,
          "author_name": "Hiram Coria 🧬",
          "author_url": "",
          "post_date": "2021-01-31T01:33:28.077000",
          "content": "<p>Thanks for share</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1177674,
      "author_name": "Adnan Zaidi",
      "author_url": "",
      "post_date": "2021-01-30T12:58:26.163000",
      "content": "<p>Superb!!! It's really a considerable job and Keep it up. </p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1176317": "![](https://media-exp1.licdn.com/dms/image/C4E22AQGgQ--ueNtBSg/feedshare-shrink_1280/0/1611887346797?e=1614816000&v=beta&t=L-D-uIphcEW35H5iB5tauxmQYzh2H8cU8w8PBdugk6c)\n\n# paper : https://arxiv.org/pdf/2101.11605.pdf\n\n# code 1 : https://github.com/leaderj1001/BottleneckTransformers\n\n# code 2 : https://github.com/lucidrains/bottleneck-transformer-pytorch",
    "1178367": "Google research and the code was made in PyTorch, interesting. ",
    "1177674": "Superb!!! It's really a considerable job and Keep it up. "
  }
}