{
  "id": 157273,
  "title": "ResourceExhaustedError",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/157273",
  "author_name": "",
  "post_date": "2020-06-10T02:01:40.924517900Z",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>ResourceExhaustedError: Failed to allocate request for 50.00GiB (53687091200B) on device ordinal 0.</p>\n\n<p>For those who successfully run image augmentation training model, how can you solve the \"resource exhausted error\" problem?</p>\n\n<p>Your response will be high appreciated.</p>\n\n<p>Thanks.</p>",
  "messages": [
    {
      "id": "880082",
      "postDate": "06/10/2020 02:01:40",
      "content": "<p>ResourceExhaustedError: Failed to allocate request for 50.00GiB (53687091200B) on device ordinal 0.</p>\n\n<p>For those who successfully run image augmentation training model, how can you solve the \"resource exhausted error\" problem?</p>\n\n<p>Your response will be high appreciated.</p>\n\n<p>Thanks.</p>",
      "rawMarkdown": "ResourceExhaustedError: Failed to allocate request for 50.00GiB (53687091200B) on device ordinal 0.\n\nFor those who successfully run image augmentation training model, how can you solve the \"resource exhausted error\" problem?\n\nYour response will be high appreciated.\n\nThanks.",
      "votes": null
    },
    {
      "id": "880102",
      "postDate": "06/10/2020 02:30:28",
      "content": "<p>It's hard to debug without seeing your code, but if you decrease your batch size that may help.</p>",
      "rawMarkdown": "It's hard to debug without seeing your code, but if you decrease your batch size that may help.",
      "votes": null
    },
    {
      "id": "880131",
      "postDate": "06/10/2020 02:59:18",
      "content": "<p>I am facing same error.</p>",
      "rawMarkdown": "I am facing same error.",
      "votes": null
    },
    {
      "id": "880133",
      "postDate": "06/10/2020 03:06:47",
      "content": "<p>There is no issue of batch size because it gives me error in model training. I have reduce the tensor size but it still gives me error. And before 2 hours this was working correctly and I have just change activation function to improve the model performance and got this error.</p>",
      "rawMarkdown": "There is no issue of batch size because it gives me error in model training. I have reduce the tensor size but it still gives me error. And before 2 hours this was working correctly and I have just change activation function to improve the model performance and got this error.",
      "votes": null
    },
    {
      "id": "896224",
      "postDate": "06/22/2020 01:54:55",
      "content": "<p>I got this error because of stride=1 . I am able to solve this error using stride=2.</p>",
      "rawMarkdown": "I got this error because of stride=1 . I am able to solve this error using stride=2.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 880102,
      "author_name": "cdeotte",
      "author_url": "",
      "post_date": "06/10/2020 02:30:28",
      "content": "<p>It's hard to debug without seeing your code, but if you decrease your batch size that may help.</p>",
      "votes": null,
      "replies": [
        {
          "id": 880133,
          "author_name": "margipatel",
          "author_url": "",
          "post_date": "06/10/2020 03:06:47",
          "content": "<p>There is no issue of batch size because it gives me error in model training. I have reduce the tensor size but it still gives me error. And before 2 hours this was working correctly and I have just change activation function to improve the model performance and got this error.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 880131,
      "author_name": "sanketpatel108784",
      "author_url": "",
      "post_date": "06/10/2020 02:59:18",
      "content": "<p>I am facing same error.</p>",
      "votes": null,
      "replies": [
        {
          "id": 896224,
          "author_name": "margipatel",
          "author_url": "",
          "post_date": "06/22/2020 01:54:55",
          "content": "<p>I got this error because of stride=1 . I am able to solve this error using stride=2.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "880082": "ResourceExhaustedError: Failed to allocate request for 50.00GiB (53687091200B) on device ordinal 0.\n\nFor those who successfully run image augmentation training model, how can you solve the \"resource exhausted error\" problem?\n\nYour response will be high appreciated.\n\nThanks.",
    "880102": "It's hard to debug without seeing your code, but if you decrease your batch size that may help.",
    "880131": "I am facing same error.",
    "880133": "There is no issue of batch size because it gives me error in model training. I have reduce the tensor size but it still gives me error. And before 2 hours this was working correctly and I have just change activation function to improve the model performance and got this error.",
    "896224": "I got this error because of stride=1 . I am able to solve this error using stride=2."
  },
  "source": "meta"
}