{
  "id": 176175,
  "title": "How do you monitor training?",
  "url": "/competitions/birdsong-recognition/discussion/176175",
  "author_name": "",
  "post_date": "2020-08-20T19:02:10.443591700Z",
  "votes": -1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi! This is somewhat of a noobish question when it comes to Kaggle kernels, but I was wondering how do you usually monitor the training process on Kaggle kernels? </p>\n<p>Do you just log the loss in a text file and use it to plot later? Or is some kind of tensorboard-ish workaround available with Kaggle kernels as well?</p>\n<p>I am using Pytorch for this competition.</p>\n<p>Thanks!</p>",
  "messages": [
    {
      "id": "979323",
      "postDate": "08/20/2020 19:02:10",
      "content": "<p>Hi! This is somewhat of a noobish question when it comes to Kaggle kernels, but I was wondering how do you usually monitor the training process on Kaggle kernels? </p>\n<p>Do you just log the loss in a text file and use it to plot later? Or is some kind of tensorboard-ish workaround available with Kaggle kernels as well?</p>\n<p>I am using Pytorch for this competition.</p>\n<p>Thanks!</p>",
      "rawMarkdown": "Hi! This is somewhat of a noobish question when it comes to Kaggle kernels, but I was wondering how do you usually monitor the training process on Kaggle kernels? \n\nDo you just log the loss in a text file and use it to plot later? Or is some kind of tensorboard-ish workaround available with Kaggle kernels as well?\n\nI am using Pytorch for this competition.\n\nThanks!",
      "votes": null
    },
    {
      "id": "979484",
      "postDate": "08/20/2020 22:00:44",
      "content": "<p>I simply print my monitoring metrics at the end of each epoch. You can obviously get more fancy--like saving your metrics into a dataframe. I know one user who set up a way to email their metrics after every epoch from Kaggle Kernels to his email.</p>\n<p>When your Notebook finishes, you can see everything that is printed.</p>",
      "rawMarkdown": "I simply print my monitoring metrics at the end of each epoch. You can obviously get more fancy--like saving your metrics into a dataframe. I know one user who set up a way to email their metrics after every epoch from Kaggle Kernels to his email.\n\nWhen your Notebook finishes, you can see everything that is printed.",
      "votes": null
    },
    {
      "id": "979846",
      "postDate": "08/21/2020 06:38:30",
      "content": "<p>Thanks! yeaah, I can track the metrics, but plotting it interactively on the graph is relatively easier to comprehend (I hate scrolling down multiple epochs)</p>\n<p>I just wanted to know if someone is actively doing something tensorboard-ish on Kaggle kernels or not. Nevermind. Thanks!</p>",
      "rawMarkdown": "Thanks! yeaah, I can track the metrics, but plotting it interactively on the graph is relatively easier to comprehend (I hate scrolling down multiple epochs)\n\nI just wanted to know if someone is actively doing something tensorboard-ish on Kaggle kernels or not. Nevermind. Thanks!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 979484,
      "author_name": "returnofsputnik",
      "author_url": "",
      "post_date": "08/20/2020 22:00:44",
      "content": "<p>I simply print my monitoring metrics at the end of each epoch. You can obviously get more fancy--like saving your metrics into a dataframe. I know one user who set up a way to email their metrics after every epoch from Kaggle Kernels to his email.</p>\n<p>When your Notebook finishes, you can see everything that is printed.</p>",
      "votes": null,
      "replies": [
        {
          "id": 979846,
          "author_name": "saran95",
          "author_url": "",
          "post_date": "08/21/2020 06:38:30",
          "content": "<p>Thanks! yeaah, I can track the metrics, but plotting it interactively on the graph is relatively easier to comprehend (I hate scrolling down multiple epochs)</p>\n<p>I just wanted to know if someone is actively doing something tensorboard-ish on Kaggle kernels or not. Nevermind. Thanks!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "979323": "Hi! This is somewhat of a noobish question when it comes to Kaggle kernels, but I was wondering how do you usually monitor the training process on Kaggle kernels? \n\nDo you just log the loss in a text file and use it to plot later? Or is some kind of tensorboard-ish workaround available with Kaggle kernels as well?\n\nI am using Pytorch for this competition.\n\nThanks!",
    "979484": "I simply print my monitoring metrics at the end of each epoch. You can obviously get more fancy--like saving your metrics into a dataframe. I know one user who set up a way to email their metrics after every epoch from Kaggle Kernels to his email.\n\nWhen your Notebook finishes, you can see everything that is printed.",
    "979846": "Thanks! yeaah, I can track the metrics, but plotting it interactively on the graph is relatively easier to comprehend (I hate scrolling down multiple epochs)\n\nI just wanted to know if someone is actively doing something tensorboard-ish on Kaggle kernels or not. Nevermind. Thanks!"
  },
  "source": "meta"
}