{
  "id": 419253,
  "title": "SMP for this competition",
  "url": "/competitions/hubmap-hacking-the-human-vasculature/discussion/419253",
  "author_name": "",
  "post_date": "2023-06-25T01:49:33.890426600Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hi, Has anyone used smp for this competition?, I think that smp models only return mask and not mask+probability.</p>\n<p>Thanks in advance!</p>",
  "messages": [
    {
      "id": "2316475",
      "postDate": "06/25/2023 01:49:33",
      "content": "<p>Hi, Has anyone used smp for this competition?, I think that smp models only return mask and not mask+probability.</p>\n<p>Thanks in advance!</p>",
      "rawMarkdown": "Hi, Has anyone used smp for this competition?, I think that smp models only return mask and not mask+probability.\n\nThanks in advance!",
      "votes": null
    },
    {
      "id": "2316559",
      "postDate": "06/25/2023 05:00:45",
      "content": "<p>Semantic segmentation models in SMP library use conv heads so their outputs are (batch_size, n_class, height, width) sized logits.<br>\n<a href=\"https://github.com/qubvel/segmentation_models.pytorch/blob/master/segmentation_models_pytorch/base/heads.py#L5-L10\" target=\"_blank\">https://github.com/qubvel/segmentation_models.pytorch/blob/master/segmentation_models_pytorch/base/heads.py#L5-L10</a></p>",
      "rawMarkdown": "Semantic segmentation models in SMP library use conv heads so their outputs are (batch_size, n_class, height, width) sized logits.\nhttps://github.com/qubvel/segmentation_models.pytorch/blob/master/segmentation_models_pytorch/base/heads.py#L5-L10",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2316559,
      "author_name": "gunesevitan",
      "author_url": "",
      "post_date": "06/25/2023 05:00:45",
      "content": "<p>Semantic segmentation models in SMP library use conv heads so their outputs are (batch_size, n_class, height, width) sized logits.<br>\n<a href=\"https://github.com/qubvel/segmentation_models.pytorch/blob/master/segmentation_models_pytorch/base/heads.py#L5-L10\" target=\"_blank\">https://github.com/qubvel/segmentation_models.pytorch/blob/master/segmentation_models_pytorch/base/heads.py#L5-L10</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2316475": "Hi, Has anyone used smp for this competition?, I think that smp models only return mask and not mask+probability.\n\nThanks in advance!",
    "2316559": "Semantic segmentation models in SMP library use conv heads so their outputs are (batch_size, n_class, height, width) sized logits.\nhttps://github.com/qubvel/segmentation_models.pytorch/blob/master/segmentation_models_pytorch/base/heads.py#L5-L10"
  },
  "source": "meta"
}