{
  "id": 134122,
  "title": "check accuracy on future epochs",
  "url": "/competitions/deepfake-detection-challenge/discussion/134122",
  "author_name": "",
  "post_date": "2020-03-06T05:16:45.868930600Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I am using fit_generator for over 32000 steps per epoch my model accuracy stabled around after 10000 steps on 1st epoch.\nthis might be a dumb question but is there a way,being a newbie i am asking</p>\n\n<p>how can i check if i can expect any increase on further epochs ? </p>",
  "messages": [
    {
      "id": "764937",
      "postDate": "03/06/2020 05:16:45",
      "content": "<p>I am using fit_generator for over 32000 steps per epoch my model accuracy stabled around after 10000 steps on 1st epoch.\nthis might be a dumb question but is there a way,being a newbie i am asking</p>\n\n<p>how can i check if i can expect any increase on further epochs ? </p>",
      "rawMarkdown": "I am using fit_generator for over 32000 steps per epoch my model accuracy stabled around after 10000 steps on 1st epoch.\nthis might be a dumb question but is there a way,being a newbie i am asking\n\nhow can i check if i can expect any increase on further epochs ?",
      "votes": null
    },
    {
      "id": "766005",
      "postDate": "03/07/2020 14:29:06",
      "content": "<p>I think it's a good question,I have ever met the same questions.\nVal-accuracy,val-loss,loss,accuracy and recall rate and so on ,\nThose values of changing can be judged by you to predict the tread of some increasing in some epochs or steps</p>",
      "rawMarkdown": "I think it's a good question,I have ever met the same questions.\nVal-accuracy,val-loss,loss,accuracy and recall rate and so on ,\nThose values of changing can be judged by you to predict the tread of some increasing in some epochs or steps",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 766005,
      "author_name": "jeffzhenyuqiu",
      "author_url": "",
      "post_date": "03/07/2020 14:29:06",
      "content": "<p>I think it's a good question,I have ever met the same questions.\nVal-accuracy,val-loss,loss,accuracy and recall rate and so on ,\nThose values of changing can be judged by you to predict the tread of some increasing in some epochs or steps</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "764937": "I am using fit_generator for over 32000 steps per epoch my model accuracy stabled around after 10000 steps on 1st epoch.\nthis might be a dumb question but is there a way,being a newbie i am asking\n\nhow can i check if i can expect any increase on further epochs ?",
    "766005": "I think it's a good question,I have ever met the same questions.\nVal-accuracy,val-loss,loss,accuracy and recall rate and so on ,\nThose values of changing can be judged by you to predict the tread of some increasing in some epochs or steps"
  },
  "source": "meta"
}