{
  "id": 147151,
  "title": "Submitting after competition end.",
  "url": "/competitions/deepfake-detection-challenge/discussion/147151",
  "author_name": "JamesRick",
  "post_date": "2020-04-29T16:58:05.163000",
  "votes": 4,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I know that submissions were closed since the resources needed to be used for evaluating models on the private leaderboard, but will submissions re-open for the public test set once all evaluations are complete? Or will the labels for the public test set be released?</p>\n\n<p>I am curious to see how well my model's performance on the cross-validation set I have matches the performance on the public test set.</p>",
  "messages": [
    {
      "id": 826468,
      "postDate": "2020-04-29T16:58:05.163Z",
      "content": "<p>I know that submissions were closed since the resources needed to be used for evaluating models on the private leaderboard, but will submissions re-open for the public test set once all evaluations are complete? Or will the labels for the public test set be released?</p>\n\n<p>I am curious to see how well my model's performance on the cross-validation set I have matches the performance on the public test set.</p>",
      "rawMarkdown": "I know that submissions were closed since the resources needed to be used for evaluating models on the private leaderboard, but will submissions re-open for the public test set once all evaluations are complete? Or will the labels for the public test set be released?\n\nI am curious to see how well my model's performance on the cross-validation set I have matches the performance on the public test set.",
      "votes": 4
    },
    {
      "id": 828440,
      "postDate": "2020-05-01T02:57:51.080Z",
      "content": "<p>Looks like many people are waiting for late submission. For sure that the late submission will be reopened, just don't know when, keep waiting and waiting. </p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2F08260a9ba9d207398680449b6812ac5f%2FScreen%20Shot%202020-05-01%20at%2012.46.38%20PM.png?generation=1588301244552934&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Ffad0217cb2b51adaa010eadd2335c778%2FScreen%20Shot%202020-05-01%20at%2012.46.57%20PM.png?generation=1588301265996241&amp;alt=media\" alt=\"\"></p>\n\n<p>As for the labels for the public test set, according to @Moshel suggestion, I found 290 labels from training set (<a href=\"https://www.kaggle.com/muerbingsha/dfdc-test-label\">https://www.kaggle.com/muerbingsha/dfdc-test-label</a>). And the correlation between these 290 test set and LB is as follows:</p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Fbf1f1ad745d450245b14f5d8820ade6f%2FScreen%20Shot%202020-05-01%20at%2012.54.03%20PM.png?generation=1588301672359159&amp;alt=media\" alt=\"\"></p>\n\n<p>My suspicion is that if the score on 290 test down to 0.0X zone, then can be confident to submit.</p>",
      "rawMarkdown": "Looks like many people are waiting for late submission. For sure that the late submission will be reopened, just don't know when, keep waiting and waiting. \n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2F08260a9ba9d207398680449b6812ac5f%2FScreen%20Shot%202020-05-01%20at%2012.46.38%20PM.png?generation=1588301244552934&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Ffad0217cb2b51adaa010eadd2335c778%2FScreen%20Shot%202020-05-01%20at%2012.46.57%20PM.png?generation=1588301265996241&amp;alt=media)\n\nAs for the labels for the public test set, according to @Moshel suggestion, I found 290 labels from training set (https://www.kaggle.com/muerbingsha/dfdc-test-label). And the correlation between these 290 test set and LB is as follows:\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Fbf1f1ad745d450245b14f5d8820ade6f%2FScreen%20Shot%202020-05-01%20at%2012.54.03%20PM.png?generation=1588301672359159&amp;alt=media)\n\nMy suspicion is that if the score on 290 test down to 0.0X zone, then can be confident to submit.\n",
      "votes": 1,
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 828440,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-05-01T02:57:51.080000",
      "content": "<p>Looks like many people are waiting for late submission. For sure that the late submission will be reopened, just don't know when, keep waiting and waiting. </p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2F08260a9ba9d207398680449b6812ac5f%2FScreen%20Shot%202020-05-01%20at%2012.46.38%20PM.png?generation=1588301244552934&amp;alt=media\" alt=\"\"></p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Ffad0217cb2b51adaa010eadd2335c778%2FScreen%20Shot%202020-05-01%20at%2012.46.57%20PM.png?generation=1588301265996241&amp;alt=media\" alt=\"\"></p>\n\n<p>As for the labels for the public test set, according to @Moshel suggestion, I found 290 labels from training set (<a href=\"https://www.kaggle.com/muerbingsha/dfdc-test-label\">https://www.kaggle.com/muerbingsha/dfdc-test-label</a>). And the correlation between these 290 test set and LB is as follows:</p>\n\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Fbf1f1ad745d450245b14f5d8820ade6f%2FScreen%20Shot%202020-05-01%20at%2012.54.03%20PM.png?generation=1588301672359159&amp;alt=media\" alt=\"\"></p>\n\n<p>My suspicion is that if the score on 290 test down to 0.0X zone, then can be confident to submit.</p>",
      "votes": 1,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "826468": "I know that submissions were closed since the resources needed to be used for evaluating models on the private leaderboard, but will submissions re-open for the public test set once all evaluations are complete? Or will the labels for the public test set be released?\n\nI am curious to see how well my model's performance on the cross-validation set I have matches the performance on the public test set.",
    "828440": "Looks like many people are waiting for late submission. For sure that the late submission will be reopened, just don't know when, keep waiting and waiting. \n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2F08260a9ba9d207398680449b6812ac5f%2FScreen%20Shot%202020-05-01%20at%2012.46.38%20PM.png?generation=1588301244552934&amp;alt=media)\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Ffad0217cb2b51adaa010eadd2335c778%2FScreen%20Shot%202020-05-01%20at%2012.46.57%20PM.png?generation=1588301265996241&amp;alt=media)\n\nAs for the labels for the public test set, according to @Moshel suggestion, I found 290 labels from training set (https://www.kaggle.com/muerbingsha/dfdc-test-label). And the correlation between these 290 test set and LB is as follows:\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2663988%2Fbf1f1ad745d450245b14f5d8820ade6f%2FScreen%20Shot%202020-05-01%20at%2012.54.03%20PM.png?generation=1588301672359159&amp;alt=media)\n\nMy suspicion is that if the score on 290 test down to 0.0X zone, then can be confident to submit.\n"
  }
}