{
  "id": 307260,
  "title": "Vote if you have used Fasterrcnn. And comment your acc",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/307260",
  "author_name": "",
  "post_date": "2022-02-13T13:58:00.025965700Z",
  "votes": -10,
  "comment_count": 8,
  "views": 0,
  "content": "<p>Vote if you have used Fasterrcnn in tenserflow and also comment on the highest accuracy that you got.</p>",
  "messages": [
    {
      "id": "1688225",
      "postDate": "02/13/2022 13:58:00",
      "content": "<p>Vote if you have used Fasterrcnn in tenserflow and also comment on the highest accuracy that you got.</p>",
      "rawMarkdown": "Vote if you have used Fasterrcnn in tenserflow and also comment on the highest accuracy that you got.",
      "votes": null
    },
    {
      "id": "1688459",
      "postDate": "02/13/2022 16:28:05",
      "content": "<p>Post a comment of your accuracy</p>",
      "rawMarkdown": "Post a comment of your accuracy",
      "votes": null
    },
    {
      "id": "1688998",
      "postDate": "02/14/2022 00:42:25",
      "content": "<p>I only used it early on in the competition as a baseline, no image augmentations or training tricks etc. best LB acc was 0.481 w/ FasterRCNN ResNet50 Backbone</p>",
      "rawMarkdown": "I only used it early on in the competition as a baseline, no image augmentations or training tricks etc. best LB acc was 0.481 w/ FasterRCNN ResNet50 Backbone",
      "votes": null
    },
    {
      "id": "1689006",
      "postDate": "02/14/2022 00:55:14",
      "content": "<p>I got 0.65+ using faster-rcnn resnet50</p>",
      "rawMarkdown": "I got 0.65+ using faster-rcnn resnet50",
      "votes": null
    },
    {
      "id": "1689109",
      "postDate": "02/14/2022 03:35:34",
      "content": "<p>Could you guys please share the notebook.</p>",
      "rawMarkdown": "Could you guys please share the notebook.",
      "votes": null
    },
    {
      "id": "1689112",
      "postDate": "02/14/2022 03:40:11",
      "content": "<p>High LB doesn't make any sense, faster-rcnn is weak on our CV.</p>",
      "rawMarkdown": "High LB doesn't make any sense, faster-rcnn is weak on our CV.",
      "votes": null
    },
    {
      "id": "1689119",
      "postDate": "02/14/2022 03:49:55",
      "content": "<p>No, sir my aim is not to be high on leader board. If that was the case I have gone with yolo v5. But aim is to experiment with it. But I have done something wrong with my model. Which is giving an acc of just 0.05 so I am making some mistake. So to know that I wanted to refer the code of others. So please 😊.</p>",
      "rawMarkdown": "No, sir my aim is not to be high on leader board. If that was the case I have gone with yolo v5. But aim is to experiment with it. But I have done something wrong with my model. Which is giving an acc of just 0.05 so I am making some mistake. So to know that I wanted to refer the code of others. So please 😊.",
      "votes": null
    },
    {
      "id": "1689157",
      "postDate": "02/14/2022 04:17:26",
      "content": "<blockquote>\n  <p>if you have used Fasterrcnn in tenserflow</p>\n</blockquote>\n<p>We used it in pytorch, so you can learn from the work of others, such as <a href=\"https://www.kaggle.com/ravishah1/cots-faster-rcnn-training-w-tf-2-0-od-api-0-474\" target=\"_blank\">https://www.kaggle.com/ravishah1/cots-faster-rcnn-training-w-tf-2-0-od-api-0-474</a></p>",
      "rawMarkdown": "> if you have used Fasterrcnn in tenserflow\n\nWe used it in pytorch, so you can learn from the work of others, such as https://www.kaggle.com/ravishah1/cots-faster-rcnn-training-w-tf-2-0-od-api-0-474",
      "votes": null
    },
    {
      "id": "1689182",
      "postDate": "02/14/2022 04:30:46",
      "content": "<p>Thank you so much</p>",
      "rawMarkdown": "Thank you so much",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1688459,
      "author_name": "captainabhijeeth",
      "author_url": "",
      "post_date": "02/13/2022 16:28:05",
      "content": "<p>Post a comment of your accuracy</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1688998,
      "author_name": "maxvandijck",
      "author_url": "",
      "post_date": "02/14/2022 00:42:25",
      "content": "<p>I only used it early on in the competition as a baseline, no image augmentations or training tricks etc. best LB acc was 0.481 w/ FasterRCNN ResNet50 Backbone</p>",
      "votes": null,
      "replies": [
        {
          "id": 1689006,
          "author_name": "klawensliu",
          "author_url": "",
          "post_date": "02/14/2022 00:55:14",
          "content": "<p>I got 0.65+ using faster-rcnn resnet50</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1689109,
          "author_name": "captainabhijeeth",
          "author_url": "",
          "post_date": "02/14/2022 03:35:34",
          "content": "<p>Could you guys please share the notebook.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1689112,
          "author_name": "klawensliu",
          "author_url": "",
          "post_date": "02/14/2022 03:40:11",
          "content": "<p>High LB doesn't make any sense, faster-rcnn is weak on our CV.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1689119,
          "author_name": "captainabhijeeth",
          "author_url": "",
          "post_date": "02/14/2022 03:49:55",
          "content": "<p>No, sir my aim is not to be high on leader board. If that was the case I have gone with yolo v5. But aim is to experiment with it. But I have done something wrong with my model. Which is giving an acc of just 0.05 so I am making some mistake. So to know that I wanted to refer the code of others. So please 😊.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1689157,
          "author_name": "zzy990106",
          "author_url": "",
          "post_date": "02/14/2022 04:17:26",
          "content": "<blockquote>\n  <p>if you have used Fasterrcnn in tenserflow</p>\n</blockquote>\n<p>We used it in pytorch, so you can learn from the work of others, such as <a href=\"https://www.kaggle.com/ravishah1/cots-faster-rcnn-training-w-tf-2-0-od-api-0-474\" target=\"_blank\">https://www.kaggle.com/ravishah1/cots-faster-rcnn-training-w-tf-2-0-od-api-0-474</a></p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1689182,
          "author_name": "captainabhijeeth",
          "author_url": "",
          "post_date": "02/14/2022 04:30:46",
          "content": "<p>Thank you so much</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1688225": "Vote if you have used Fasterrcnn in tenserflow and also comment on the highest accuracy that you got.",
    "1688459": "Post a comment of your accuracy",
    "1688998": "I only used it early on in the competition as a baseline, no image augmentations or training tricks etc. best LB acc was 0.481 w/ FasterRCNN ResNet50 Backbone",
    "1689006": "I got 0.65+ using faster-rcnn resnet50",
    "1689109": "Could you guys please share the notebook.",
    "1689112": "High LB doesn't make any sense, faster-rcnn is weak on our CV.",
    "1689119": "No, sir my aim is not to be high on leader board. If that was the case I have gone with yolo v5. But aim is to experiment with it. But I have done something wrong with my model. Which is giving an acc of just 0.05 so I am making some mistake. So to know that I wanted to refer the code of others. So please 😊.",
    "1689157": "> if you have used Fasterrcnn in tenserflow\n\nWe used it in pytorch, so you can learn from the work of others, such as https://www.kaggle.com/ravishah1/cots-faster-rcnn-training-w-tf-2-0-od-api-0-474",
    "1689182": "Thank you so much"
  },
  "source": "meta"
}