{
  "id": 96940,
  "title": "Cluster-GCN: cool and fast training of GCN",
  "url": "/competitions/champs-scalar-coupling/discussion/96940",
  "author_name": "Andrey Lukyanenko",
  "post_date": "2019-06-24T17:45:29.700000",
  "votes": 16,
  "comment_count": 4,
  "views": 0,
  "content": "<p><a href=\"https://www.reddit.com/r/MachineLearning/comments/c4kkux/r_a_pytorch_implementation_of_clustergcn_an/\">https://www.reddit.com/r/MachineLearning/comments/c4kkux/r_a_pytorch_implementation_of_clustergcn_an/</a>\nA PyTorch implementation of \"Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks\"</p>\n\n<p>For training a 3-layer GCN on this data, Cluster-GCN is faster than the previous state-of-the-art VR-GCN (1523 seconds vs 1961 seconds) and using much less memory (2.2GB vs 11.2GB). Furthermore, for training 4 layer GCN on this data, our algorithm can finish in around 36 minutes while all the existing GCN training algorithms fail to train due to the out-of-memory issue.</p>",
  "messages": [
    {
      "id": 559865,
      "postDate": "2019-06-24T17:45:29.700Z",
      "content": "<p><a href=\"https://www.reddit.com/r/MachineLearning/comments/c4kkux/r_a_pytorch_implementation_of_clustergcn_an/\">https://www.reddit.com/r/MachineLearning/comments/c4kkux/r_a_pytorch_implementation_of_clustergcn_an/</a>\nA PyTorch implementation of \"Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks\"</p>\n\n<p>For training a 3-layer GCN on this data, Cluster-GCN is faster than the previous state-of-the-art VR-GCN (1523 seconds vs 1961 seconds) and using much less memory (2.2GB vs 11.2GB). Furthermore, for training 4 layer GCN on this data, our algorithm can finish in around 36 minutes while all the existing GCN training algorithms fail to train due to the out-of-memory issue.</p>",
      "rawMarkdown": "https://www.reddit.com/r/MachineLearning/comments/c4kkux/r_a_pytorch_implementation_of_clustergcn_an/\nA PyTorch implementation of \"Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks\"\n\nFor training a 3-layer GCN on this data, Cluster-GCN is faster than the previous state-of-the-art VR-GCN (1523 seconds vs 1961 seconds) and using much less memory (2.2GB vs 11.2GB). Furthermore, for training 4 layer GCN on this data, our algorithm can finish in around 36 minutes while all the existing GCN training algorithms fail to train due to the out-of-memory issue.",
      "votes": 16
    },
    {
      "id": 563538,
      "postDate": "2019-06-28T11:52:10.747Z",
      "content": "<p>Just wondering, will this help much for this challenge, as this seems to aim to speed up GCNs for huge graphs, where we are dealing with max 29 nodes?</p>",
      "rawMarkdown": "Just wondering, will this help much for this challenge, as this seems to aim to speed up GCNs for huge graphs, where we are dealing with max 29 nodes?"
    },
    {
      "id": 559911,
      "postDate": "2019-06-24T19:03:36.797Z",
      "content": "<p>can you redirect me to a tensorflow implementation of this(if possible)?</p>",
      "rawMarkdown": "can you redirect me to a tensorflow implementation of this(if possible)?",
      "replies": [
        {
          "id": 559912,
          "postDate": "2019-06-24T19:05:37.513Z",
          "content": "<p>Currently there is only Pytorch implementation.</p>",
          "rawMarkdown": "Currently there is only Pytorch implementation.",
          "votes": 1
        }
      ]
    },
    {
      "id": 2132675,
      "postDate": "2023-02-07T01:34:41.047Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 563538,
      "author_name": "fnands",
      "author_url": "",
      "post_date": "2019-06-28T11:52:10.747000",
      "content": "<p>Just wondering, will this help much for this challenge, as this seems to aim to speed up GCNs for huge graphs, where we are dealing with max 29 nodes?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 559911,
      "author_name": "Mobassir",
      "author_url": "",
      "post_date": "2019-06-24T19:03:36.797000",
      "content": "<p>can you redirect me to a tensorflow implementation of this(if possible)?</p>",
      "votes": 0,
      "replies": [
        {
          "id": 559912,
          "author_name": "Andrey Lukyanenko",
          "author_url": "",
          "post_date": "2019-06-24T19:05:37.513000",
          "content": "<p>Currently there is only Pytorch implementation.</p>",
          "votes": 1,
          "replies": []
        }
      ]
    },
    {
      "id": 2132675,
      "author_name": "",
      "author_url": "",
      "post_date": "2023-02-07T01:34:41.047000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "559865": "https://www.reddit.com/r/MachineLearning/comments/c4kkux/r_a_pytorch_implementation_of_clustergcn_an/\nA PyTorch implementation of \"Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks\"\n\nFor training a 3-layer GCN on this data, Cluster-GCN is faster than the previous state-of-the-art VR-GCN (1523 seconds vs 1961 seconds) and using much less memory (2.2GB vs 11.2GB). Furthermore, for training 4 layer GCN on this data, our algorithm can finish in around 36 minutes while all the existing GCN training algorithms fail to train due to the out-of-memory issue.",
    "563538": "Just wondering, will this help much for this challenge, as this seems to aim to speed up GCNs for huge graphs, where we are dealing with max 29 nodes?",
    "559911": "can you redirect me to a tensorflow implementation of this(if possible)?",
    "2132675": ""
  }
}