{
  "id": 143746,
  "title": "Single best model",
  "url": "/competitions/flower-classification-with-tpus/discussion/143746",
  "author_name": "",
  "post_date": "2020-04-16T05:18:22.866697900Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>What is your best score for a single model with no ensembling?</p>",
  "messages": [
    {
      "id": "809356",
      "postDate": "04/16/2020 05:18:22",
      "content": "<p>What is your best score for a single model with no ensembling?</p>",
      "rawMarkdown": "What is your best score for a single model with no ensembling?",
      "votes": null
    },
    {
      "id": "811545",
      "postDate": "04/18/2020 03:50:32",
      "content": "<p>Original + <a href=\"https://www.kaggle.com/szacho/oxford-102-for-tpu-competition\">oxford</a> dataset\nEfficientNetB7\nLB: 0.96602\nCV: 0.9678\nwith TTA</p>",
      "rawMarkdown": "Original + [oxford](https://www.kaggle.com/szacho/oxford-102-for-tpu-competition) dataset\nEfficientNetB7\nLB: 0.96602\nCV: 0.9678\nwith TTA",
      "votes": null
    },
    {
      "id": "823911",
      "postDate": "04/28/2020 02:28:39",
      "content": "<p>External data</p>\n\n<p>EfficientNetB7+mixup\nLB: 0.95931\nCV: 0.99(i calculated valid_f1score)</p>",
      "rawMarkdown": "External data\n\nEfficientNetB7+mixup\nLB: 0.95931\nCV: 0.99(i calculated valid_f1score)",
      "votes": null
    },
    {
      "id": "832269",
      "postDate": "05/04/2020 02:30:04",
      "content": "<p>External data</p>\n\n<p>EfficientNetB7 w/ cutmix + mixup + some augs\nLB : 0.97573\nCV : 0.9531 (valid f1 score)\nw/o TTA</p>",
      "rawMarkdown": "External data\n\nEfficientNetB7 w/ cutmix + mixup + some augs\nLB : 0.97573\nCV : 0.9531 (valid f1 score)\nw/o TTA",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 811545,
      "author_name": "xiejialun",
      "author_url": "",
      "post_date": "04/18/2020 03:50:32",
      "content": "<p>Original + <a href=\"https://www.kaggle.com/szacho/oxford-102-for-tpu-competition\">oxford</a> dataset\nEfficientNetB7\nLB: 0.96602\nCV: 0.9678\nwith TTA</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 823911,
      "author_name": "kurianbenoy",
      "author_url": "",
      "post_date": "04/28/2020 02:28:39",
      "content": "<p>External data</p>\n\n<p>EfficientNetB7+mixup\nLB: 0.95931\nCV: 0.99(i calculated valid_f1score)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 832269,
      "author_name": "kozistr",
      "author_url": "",
      "post_date": "05/04/2020 02:30:04",
      "content": "<p>External data</p>\n\n<p>EfficientNetB7 w/ cutmix + mixup + some augs\nLB : 0.97573\nCV : 0.9531 (valid f1 score)\nw/o TTA</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "809356": "What is your best score for a single model with no ensembling?",
    "811545": "Original + [oxford](https://www.kaggle.com/szacho/oxford-102-for-tpu-competition) dataset\nEfficientNetB7\nLB: 0.96602\nCV: 0.9678\nwith TTA",
    "823911": "External data\n\nEfficientNetB7+mixup\nLB: 0.95931\nCV: 0.99(i calculated valid_f1score)",
    "832269": "External data\n\nEfficientNetB7 w/ cutmix + mixup + some augs\nLB : 0.97573\nCV : 0.9531 (valid f1 score)\nw/o TTA"
  },
  "source": "meta"
}