{
  "id": 105266,
  "title": "Kernel died always while training",
  "url": "/competitions/aptos2019-blindness-detection/discussion/105266",
  "author_name": "",
  "post_date": "2019-08-22T04:42:22.412177200Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I am using EfficientNet B0 with input size ( 224,224,3). My kernel is always died just before running a single epoch.\nCan you help me, I don't understand why this is happening??\nThanks in advance</p>",
  "messages": [
    {
      "id": "605112",
      "postDate": "08/22/2019 04:42:22",
      "content": "<p>I am using EfficientNet B0 with input size ( 224,224,3). My kernel is always died just before running a single epoch.\nCan you help me, I don't understand why this is happening??\nThanks in advance</p>",
      "rawMarkdown": "I am using EfficientNet B0 with input size ( 224,224,3). My kernel is always died just before running a single epoch.\nCan you help me, I don't understand why this is happening??\nThanks in advance",
      "votes": null
    },
    {
      "id": "605259",
      "postDate": "08/22/2019 07:49:22",
      "content": "<p>Not fully sure , but this may happen when you load everything in RAM at once so RAM Overflow may cause the kernel to die. Try reducing the data you are keeping in RAM at once.</p>",
      "rawMarkdown": "Not fully sure , but this may happen when you load everything in RAM at once so RAM Overflow may cause the kernel to die. Try reducing the data you are keeping in RAM at once.",
      "votes": null
    },
    {
      "id": "605426",
      "postDate": "08/22/2019 11:19:10",
      "content": "<p>input data when you need, you can use generator. I hope it will be helpful to you</p>",
      "rawMarkdown": "input data when you need, you can use generator. I hope it will be helpful to you",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 605259,
      "author_name": "abhijit96",
      "author_url": "",
      "post_date": "08/22/2019 07:49:22",
      "content": "<p>Not fully sure , but this may happen when you load everything in RAM at once so RAM Overflow may cause the kernel to die. Try reducing the data you are keeping in RAM at once.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 605426,
      "author_name": "jiangkun2",
      "author_url": "",
      "post_date": "08/22/2019 11:19:10",
      "content": "<p>input data when you need, you can use generator. I hope it will be helpful to you</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "605112": "I am using EfficientNet B0 with input size ( 224,224,3). My kernel is always died just before running a single epoch.\nCan you help me, I don't understand why this is happening??\nThanks in advance",
    "605259": "Not fully sure , but this may happen when you load everything in RAM at once so RAM Overflow may cause the kernel to die. Try reducing the data you are keeping in RAM at once.",
    "605426": "input data when you need, you can use generator. I hope it will be helpful to you"
  },
  "source": "meta"
}