{
  "id": 178788,
  "title": "How to use GLDv1",
  "url": "/competitions/landmark-recognition-2020/discussion/178788",
  "author_name": "",
  "post_date": "2020-08-31T12:04:38.176991300Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Does this mean using v1 as a pretraining, I don't how to train it with GLDv1 and predict witch v2  <img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3042915%2F2416b0d065b9576346b4cdd4feafbf91%2FSnipaste_2020-08-31_19-53-58.png?generation=1598874858933880&amp;alt=media\" alt=\"\"><br>\nThis is official paper of google landmark dataset<br>\nthe description of the Table: Baseline results (% µAP) for the GLDv2 recognition task</p>",
  "messages": [
    {
      "id": "992738",
      "postDate": "08/31/2020 12:04:38",
      "content": "<p>Does this mean using v1 as a pretraining, I don't how to train it with GLDv1 and predict witch v2  <img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3042915%2F2416b0d065b9576346b4cdd4feafbf91%2FSnipaste_2020-08-31_19-53-58.png?generation=1598874858933880&amp;alt=media\" alt=\"\"><br>\nThis is official paper of google landmark dataset<br>\nthe description of the Table: Baseline results (% µAP) for the GLDv2 recognition task</p>",
      "rawMarkdown": "Does this mean using v1 as a pretraining, I don't how to train it with GLDv1 and predict witch v2  ![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3042915%2F2416b0d065b9576346b4cdd4feafbf91%2FSnipaste_2020-08-31_19-53-58.png?generation=1598874858933880&alt=media)\nThis is official paper of google landmark dataset\nthe description of the Table: Baseline results (% µAP) for the GLDv2 recognition task",
      "votes": null
    },
    {
      "id": "992916",
      "postDate": "08/31/2020 14:16:00",
      "content": "<p>Good question, I had the same.</p>",
      "rawMarkdown": "Good question, I had the same.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 992916,
      "author_name": "philippsinger",
      "author_url": "",
      "post_date": "08/31/2020 14:16:00",
      "content": "<p>Good question, I had the same.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "992738": "Does this mean using v1 as a pretraining, I don't how to train it with GLDv1 and predict witch v2  ![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3042915%2F2416b0d065b9576346b4cdd4feafbf91%2FSnipaste_2020-08-31_19-53-58.png?generation=1598874858933880&alt=media)\nThis is official paper of google landmark dataset\nthe description of the Table: Baseline results (% µAP) for the GLDv2 recognition task",
    "992916": "Good question, I had the same."
  },
  "source": "meta"
}