{
  "id": 198197,
  "title": "Transfer learning - Which model best suits for cassava leaf classification ? ",
  "url": "/competitions/cassava-leaf-disease-classification/discussion/198197",
  "author_name": "Praveen ",
  "post_date": "2020-11-20T07:59:03.363000",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Which model best suits for cassava leaf classification ? </p>\n<ul>\n<li>VGGnet or any other model - Suggestion please </li>\n</ul>",
  "messages": [
    {
      "id": 1084618,
      "postDate": "2020-11-20T07:59:03.363Z",
      "content": "<p>Which model best suits for cassava leaf classification ? </p>\n<ul>\n<li>VGGnet or any other model - Suggestion please </li>\n</ul>",
      "rawMarkdown": "Which model best suits for cassava leaf classification ? \n - VGGnet or any other model - Suggestion please ",
      "votes": 1
    },
    {
      "id": 1088949,
      "postDate": "2020-11-24T04:45:59.083Z",
      "content": "<p>Start with vgg as others have suggested. But from most of the papers quoted on other posts , looks like the winning models for the last years used resnet </p>",
      "rawMarkdown": "Start with vgg as others have suggested. But from most of the papers quoted on other posts , looks like the winning models for the last years used resnet \n"
    },
    {
      "id": 1088654,
      "postDate": "2020-11-23T20:41:24.260Z",
      "content": "<p>Just try multiple different models (3-5) and compare them with each other, but make sure that you are using same hyperparameters and the overall setting should be the same. Afterwards you can see which approach works better for you. Personally i'm using EfficientNet-B1.</p>",
      "rawMarkdown": "Just try multiple different models (3-5) and compare them with each other, but make sure that you are using same hyperparameters and the overall setting should be the same. Afterwards you can see which approach works better for you. Personally i'm using EfficientNet-B1."
    },
    {
      "id": 1084649,
      "postDate": "2020-11-20T08:42:51.367Z",
      "rawMarkdown": "",
      "votes": 2,
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 1088949,
      "author_name": "venkata yerubandi",
      "author_url": "",
      "post_date": "2020-11-24T04:45:59.083000",
      "content": "<p>Start with vgg as others have suggested. But from most of the papers quoted on other posts , looks like the winning models for the last years used resnet </p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1088654,
      "author_name": "Ali Abdin",
      "author_url": "",
      "post_date": "2020-11-23T20:41:24.260000",
      "content": "<p>Just try multiple different models (3-5) and compare them with each other, but make sure that you are using same hyperparameters and the overall setting should be the same. Afterwards you can see which approach works better for you. Personally i'm using EfficientNet-B1.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1084649,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-11-20T08:42:51.367000",
      "content": "",
      "votes": 2,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1084618": "Which model best suits for cassava leaf classification ? \n - VGGnet or any other model - Suggestion please ",
    "1088949": "Start with vgg as others have suggested. But from most of the papers quoted on other posts , looks like the winning models for the last years used resnet \n",
    "1088654": "Just try multiple different models (3-5) and compare them with each other, but make sure that you are using same hyperparameters and the overall setting should be the same. Afterwards you can see which approach works better for you. Personally i'm using EfficientNet-B1.",
    "1084649": ""
  }
}