{
  "id": 302741,
  "title": "Yolov5 with higher size being timed out",
  "url": "/competitions/tensorflow-great-barrier-reef/discussion/302741",
  "author_name": "",
  "post_date": "2022-01-24T04:35:28.578063Z",
  "votes": 1,
  "comment_count": 7,
  "views": 0,
  "content": "<p>I just copied <a href=\"https://www.kaggle.com/macxiao/higher-resolution-and-confidence\" target=\"_blank\">https://www.kaggle.com/macxiao/higher-resolution-and-confidence</a> notebook and ran the inference and submitted. But the submission got timed out even though the original notebook shows a score of 0.626. </p>\n<p>Does anyone have any suggestions regarding this and in general how to use yolov5 with higher image size or resolution?</p>",
  "messages": [
    {
      "id": "1662148",
      "postDate": "01/24/2022 04:35:28",
      "content": "<p>I just copied <a href=\"https://www.kaggle.com/macxiao/higher-resolution-and-confidence\" target=\"_blank\">https://www.kaggle.com/macxiao/higher-resolution-and-confidence</a> notebook and ran the inference and submitted. But the submission got timed out even though the original notebook shows a score of 0.626. </p>\n<p>Does anyone have any suggestions regarding this and in general how to use yolov5 with higher image size or resolution?</p>",
      "rawMarkdown": "I just copied [https://www.kaggle.com/macxiao/higher-resolution-and-confidence](https://www.kaggle.com/macxiao/higher-resolution-and-confidence) notebook and ran the inference and submitted. But the submission got timed out even though the original notebook shows a score of 0.626. \n\nDoes anyone have any suggestions regarding this and in general how to use yolov5 with higher image size or resolution?",
      "votes": null
    },
    {
      "id": "1662235",
      "postDate": "01/24/2022 06:16:18",
      "content": "<p>Hi, <a href=\"https://www.kaggle.com/somnathjena\" target=\"_blank\">@somnathjena</a>. Did you use GPU or CPU? It is faster if you use GPU. I hope that helps. </p>",
      "rawMarkdown": "Hi, @somnathjena. Did you use GPU or CPU? It is faster if you use GPU. I hope that helps.",
      "votes": null
    },
    {
      "id": "1662239",
      "postDate": "01/24/2022 06:21:08",
      "content": "<p><a href=\"https://www.kaggle.com/anhnguyen811\" target=\"_blank\">@anhnguyen811</a> obviously I used GPU. Even the same code with size 9000 gave a score of 0.613.</p>",
      "rawMarkdown": "anhnguyen811 obviously I used GPU. Even the same code with size 9000 gave a score of 0.613.",
      "votes": null
    },
    {
      "id": "1662596",
      "postDate": "01/24/2022 12:12:43",
      "content": "<p>You can try running training on Colab or some other platform and run inference on Kaggle. </p>",
      "rawMarkdown": "You can try running training on Colab or some other platform and run inference on Kaggle.",
      "votes": null
    },
    {
      "id": "1662600",
      "postDate": "01/24/2022 12:16:50",
      "content": "<p><a href=\"https://www.kaggle.com/crained\" target=\"_blank\">@crained</a> we have the pretrained model in the public datasets of kaggle. I am simply running the inference only.</p>",
      "rawMarkdown": "crained we have the pretrained model in the public datasets of kaggle. I am simply running the inference only.",
      "votes": null
    },
    {
      "id": "1662626",
      "postDate": "01/24/2022 12:40:06",
      "content": "<p>Are you sure that you changed nothing in the original notebook? When I ran it it  worked under 5 hrs.</p>",
      "rawMarkdown": "Are you sure that you changed nothing in the original notebook? When I ran it it  worked under 5 hrs.",
      "votes": null
    },
    {
      "id": "1665952",
      "postDate": "01/27/2022 08:58:23",
      "content": "<p><a href=\"https://www.kaggle.com/somnathjena\" target=\"_blank\">@somnathjena</a>. Thanks for letting me know. Could you please share your error snippet here for us to understand what the error is? </p>",
      "rawMarkdown": "somnathjena. Thanks for letting me know. Could you please share your error snippet here for us to understand what the error is?",
      "votes": null
    },
    {
      "id": "1665982",
      "postDate": "01/27/2022 09:34:53",
      "content": "<p>I think you are using the latest version of this notebook, but kaggle shows the best score, achieved by notebook</p>",
      "rawMarkdown": "I think you are using the latest version of this notebook, but kaggle shows the best score, achieved by notebook",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1662235,
      "author_name": "anhnguyen811",
      "author_url": "",
      "post_date": "01/24/2022 06:16:18",
      "content": "<p>Hi, <a href=\"https://www.kaggle.com/somnathjena\" target=\"_blank\">@somnathjena</a>. Did you use GPU or CPU? It is faster if you use GPU. I hope that helps. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1662239,
      "author_name": "somnathjena",
      "author_url": "",
      "post_date": "01/24/2022 06:21:08",
      "content": "<p><a href=\"https://www.kaggle.com/anhnguyen811\" target=\"_blank\">@anhnguyen811</a> obviously I used GPU. Even the same code with size 9000 gave a score of 0.613.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1665952,
          "author_name": "anhnguyen811",
          "author_url": "",
          "post_date": "01/27/2022 08:58:23",
          "content": "<p><a href=\"https://www.kaggle.com/somnathjena\" target=\"_blank\">@somnathjena</a>. Thanks for letting me know. Could you please share your error snippet here for us to understand what the error is? </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1665982,
          "author_name": "kocha1",
          "author_url": "",
          "post_date": "01/27/2022 09:34:53",
          "content": "<p>I think you are using the latest version of this notebook, but kaggle shows the best score, achieved by notebook</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1662596,
      "author_name": "crained",
      "author_url": "",
      "post_date": "01/24/2022 12:12:43",
      "content": "<p>You can try running training on Colab or some other platform and run inference on Kaggle. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1662600,
      "author_name": "somnathjena",
      "author_url": "",
      "post_date": "01/24/2022 12:16:50",
      "content": "<p><a href=\"https://www.kaggle.com/crained\" target=\"_blank\">@crained</a> we have the pretrained model in the public datasets of kaggle. I am simply running the inference only.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1662626,
      "author_name": "elmaxuno",
      "author_url": "",
      "post_date": "01/24/2022 12:40:06",
      "content": "<p>Are you sure that you changed nothing in the original notebook? When I ran it it  worked under 5 hrs.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1662148": "I just copied [https://www.kaggle.com/macxiao/higher-resolution-and-confidence](https://www.kaggle.com/macxiao/higher-resolution-and-confidence) notebook and ran the inference and submitted. But the submission got timed out even though the original notebook shows a score of 0.626. \n\nDoes anyone have any suggestions regarding this and in general how to use yolov5 with higher image size or resolution?",
    "1662235": "Hi, @somnathjena. Did you use GPU or CPU? It is faster if you use GPU. I hope that helps.",
    "1662239": "anhnguyen811 obviously I used GPU. Even the same code with size 9000 gave a score of 0.613.",
    "1662596": "You can try running training on Colab or some other platform and run inference on Kaggle.",
    "1662600": "crained we have the pretrained model in the public datasets of kaggle. I am simply running the inference only.",
    "1662626": "Are you sure that you changed nothing in the original notebook? When I ran it it  worked under 5 hrs.",
    "1665952": "somnathjena. Thanks for letting me know. Could you please share your error snippet here for us to understand what the error is?",
    "1665982": "I think you are using the latest version of this notebook, but kaggle shows the best score, achieved by notebook"
  },
  "source": "meta"
}