{
  "id": 134589,
  "title": "how to visualize effects epoch, learning rate, optimizer ",
  "url": "/competitions/bengaliai-cv19/discussion/134589",
  "author_name": "",
  "post_date": "2020-03-09T05:24:22.491987100Z",
  "votes": 28,
  "comment_count": 8,
  "views": 0,
  "content": "<p>i find this interesting\n<a href=\"https://joshvarty.com/2019/04/16/getting-inside-a-neural-network/\">https://joshvarty.com/2019/04/16/getting-inside-a-neural-network/</a></p>\n\n<p><img src=\"https://joshvarty.files.wordpress.com/2019/04/activations2.png\" alt=\"\"></p>",
  "messages": [
    {
      "id": "767059",
      "postDate": "03/09/2020 05:24:22",
      "content": "<p>i find this interesting\n<a href=\"https://joshvarty.com/2019/04/16/getting-inside-a-neural-network/\">https://joshvarty.com/2019/04/16/getting-inside-a-neural-network/</a></p>\n\n<p><img src=\"https://joshvarty.files.wordpress.com/2019/04/activations2.png\" alt=\"\"></p>",
      "rawMarkdown": "i find this interesting\nhttps://joshvarty.com/2019/04/16/getting-inside-a-neural-network/\n\n![](https://joshvarty.files.wordpress.com/2019/04/activations2.png)",
      "votes": null
    },
    {
      "id": "767130",
      "postDate": "03/09/2020 07:50:13",
      "content": "<p>Great <a href=\"/hengck23\">@hengck23</a> thanks for sharing</p>",
      "rawMarkdown": "Great @hengck23 thanks for sharing",
      "votes": null
    },
    {
      "id": "767289",
      "postDate": "03/09/2020 12:47:45",
      "content": "<p>lovely, thanks for sharing </p>",
      "rawMarkdown": "lovely, thanks for sharing",
      "votes": null
    },
    {
      "id": "767429",
      "postDate": "03/09/2020 16:08:25",
      "content": "<p>Thanks for sharing, in this competition, you are like my window to the world 👀 </p>",
      "rawMarkdown": "Thanks for sharing, in this competition, you are like my window to the world 👀",
      "votes": null
    },
    {
      "id": "767993",
      "postDate": "03/10/2020 10:52:50",
      "content": "<p>Something related: <a href=\"https://www.wandb.com/\">wandb</a> plots the changes in weight for every layer in the model. I have no idea how to infer them, someone might. Also, wandb in general is an amazing replacement to tensorboard for me.</p>",
      "rawMarkdown": "Something related: [wandb](https://www.wandb.com/) plots the changes in weight for every layer in the model. I have no idea how to infer them, someone might. Also, wandb in general is an amazing replacement to tensorboard for me.",
      "votes": null
    },
    {
      "id": "768064",
      "postDate": "03/10/2020 12:18:00",
      "content": "<p>Awesome. Thanks for sharing!</p>",
      "rawMarkdown": "Awesome. Thanks for sharing!",
      "votes": null
    },
    {
      "id": "768084",
      "postDate": "03/10/2020 12:46:59",
      "content": "<p>Awesome. Thanks for sharing!</p>",
      "rawMarkdown": "Awesome. Thanks for sharing!",
      "votes": null
    },
    {
      "id": "768847",
      "postDate": "03/11/2020 08:55:10",
      "content": "<p>I agreed with you. Frog brother is so kind to share everything. 😄</p>",
      "rawMarkdown": "I agreed with you. Frog brother is so kind to share everything. 😄",
      "votes": null
    },
    {
      "id": "769704",
      "postDate": "03/12/2020 06:55:34",
      "content": "<p>Thanks for sharing</p>",
      "rawMarkdown": "Thanks for sharing",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 767130,
      "author_name": "chirag9073",
      "author_url": "",
      "post_date": "03/09/2020 07:50:13",
      "content": "<p>Great <a href=\"/hengck23\">@hengck23</a> thanks for sharing</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 767289,
      "author_name": "akashshingha850",
      "author_url": "",
      "post_date": "03/09/2020 12:47:45",
      "content": "<p>lovely, thanks for sharing </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 767429,
      "author_name": "barnwellguy",
      "author_url": "",
      "post_date": "03/09/2020 16:08:25",
      "content": "<p>Thanks for sharing, in this competition, you are like my window to the world 👀 </p>",
      "votes": null,
      "replies": [
        {
          "id": 768847,
          "author_name": "cnzengshiyuan",
          "author_url": "",
          "post_date": "03/11/2020 08:55:10",
          "content": "<p>I agreed with you. Frog brother is so kind to share everything. 😄</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 767993,
      "author_name": "timetraveller98",
      "author_url": "",
      "post_date": "03/10/2020 10:52:50",
      "content": "<p>Something related: <a href=\"https://www.wandb.com/\">wandb</a> plots the changes in weight for every layer in the model. I have no idea how to infer them, someone might. Also, wandb in general is an amazing replacement to tensorboard for me.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 768064,
      "author_name": "olavomendes",
      "author_url": "",
      "post_date": "03/10/2020 12:18:00",
      "content": "<p>Awesome. Thanks for sharing!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 768084,
      "author_name": "olavomendes",
      "author_url": "",
      "post_date": "03/10/2020 12:46:59",
      "content": "<p>Awesome. Thanks for sharing!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 769704,
      "author_name": "jagannathrk",
      "author_url": "",
      "post_date": "03/12/2020 06:55:34",
      "content": "<p>Thanks for sharing</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "767059": "i find this interesting\nhttps://joshvarty.com/2019/04/16/getting-inside-a-neural-network/\n\n![](https://joshvarty.files.wordpress.com/2019/04/activations2.png)",
    "767130": "Great @hengck23 thanks for sharing",
    "767289": "lovely, thanks for sharing",
    "767429": "Thanks for sharing, in this competition, you are like my window to the world 👀",
    "767993": "Something related: [wandb](https://www.wandb.com/) plots the changes in weight for every layer in the model. I have no idea how to infer them, someone might. Also, wandb in general is an amazing replacement to tensorboard for me.",
    "768064": "Awesome. Thanks for sharing!",
    "768084": "Awesome. Thanks for sharing!",
    "768847": "I agreed with you. Frog brother is so kind to share everything. 😄",
    "769704": "Thanks for sharing"
  },
  "source": "meta"
}