{
  "id": 475360,
  "title": "Late Submission results",
  "url": "/competitions/copy-of-pathologyaidemo2/discussion/475360",
  "author_name": "enigmanx",
  "post_date": "2024-02-08T05:53:05.545000",
  "votes": 2,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I tried several variations, but did not exceed the baseline code. I found forward-and-backwarding models with different learning parameters  is beneficial for reducing overhead during training (multiple learning rates variants). I am running CLAM-SB-like MIL pipeline and will share the results in a couple of days.</p>",
  "messages": [
    {
      "id": 2642364,
      "postDate": "2024-02-08T05:53:05.547Z",
      "content": "<p>I tried several variations, but did not exceed the baseline code. I found forward-and-backwarding models with different learning parameters  is beneficial for reducing overhead during training (multiple learning rates variants). I am running CLAM-SB-like MIL pipeline and will share the results in a couple of days.</p>",
      "rawMarkdown": "I tried several variations, but did not exceed the baseline code. I found forward-and-backwarding models with different learning parameters  is beneficial for reducing overhead during training (multiple learning rates variants). I am running CLAM-SB-like MIL pipeline and will share the results in a couple of days.",
      "votes": 2
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2642364": "I tried several variations, but did not exceed the baseline code. I found forward-and-backwarding models with different learning parameters  is beneficial for reducing overhead during training (multiple learning rates variants). I am running CLAM-SB-like MIL pipeline and will share the results in a couple of days."
  }
}