{
  "id": 214289,
  "title": "How to download and upload weights, Need Help !",
  "url": "/competitions/cassava-leaf-disease-classification/discussion/214289",
  "author_name": "",
  "post_date": "2021-01-26T03:37:50.736702200Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hi there, I need little bit of help, I am saving the trained weights using checkpoint callback ,how can I download and upload my trained weights ? I am using EfficientNetB0 with keras.</p>",
  "messages": [
    {
      "id": "1170184",
      "postDate": "01/26/2021 03:37:50",
      "content": "<p>Hi there, I need little bit of help, I am saving the trained weights using checkpoint callback ,how can I download and upload my trained weights ? I am using EfficientNetB0 with keras.</p>",
      "rawMarkdown": "Hi there, I need little bit of help, I am saving the trained weights using checkpoint callback ,how can I download and upload my trained weights ? I am using EfficientNetB0 with keras.",
      "votes": null
    },
    {
      "id": "1170197",
      "postDate": "01/26/2021 03:57:49",
      "content": "<p>Hello!</p>\n<p>First of all, make sure that your <code>ModelCheckpoint</code> callback actually saves your weights by checking your <code>/kaggle/working</code> directory.<br>\nThen you can hit <strong>Save Version</strong> (Save &amp; Run All (Commit)) in your notebook. By going to the <strong>Output</strong> section in the Viewer and hitting <strong>+ New Dataset</strong> you can create a new private dataset containing all the outputs of your notebook. The dataset can then be imported to your inference notebook. Or simply download the results from the <strong>Output</strong> section to use them on your local machine.</p>",
      "rawMarkdown": "Hello!\n\nFirst of all, make sure that your `ModelCheckpoint` callback actually saves your weights by checking your `/kaggle/working` directory.\nThen you can hit **Save Version** (Save & Run All (Commit)) in your notebook. By going to the **Output** section in the Viewer and hitting **+ New Dataset** you can create a new private dataset containing all the outputs of your notebook. The dataset can then be imported to your inference notebook. Or simply download the results from the **Output** section to use them on your local machine.",
      "votes": null
    },
    {
      "id": "1170198",
      "postDate": "01/26/2021 03:58:03",
      "content": "<p>Open the running notebook and click &lt; output &gt; to view the output data. There is a downward arrow on the right of download, or click download all on the left.</p>",
      "rawMarkdown": "Open the running notebook and click < output > to view the output data. There is a downward arrow on the right of download, or click download all on the left.",
      "votes": null
    },
    {
      "id": "1170295",
      "postDate": "01/26/2021 05:48:21",
      "content": "<p>thank you for the suggestion</p>",
      "rawMarkdown": "thank you for the suggestion",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1170197,
      "author_name": "nickuzmenkov",
      "author_url": "",
      "post_date": "01/26/2021 03:57:49",
      "content": "<p>Hello!</p>\n<p>First of all, make sure that your <code>ModelCheckpoint</code> callback actually saves your weights by checking your <code>/kaggle/working</code> directory.<br>\nThen you can hit <strong>Save Version</strong> (Save &amp; Run All (Commit)) in your notebook. By going to the <strong>Output</strong> section in the Viewer and hitting <strong>+ New Dataset</strong> you can create a new private dataset containing all the outputs of your notebook. The dataset can then be imported to your inference notebook. Or simply download the results from the <strong>Output</strong> section to use them on your local machine.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1170295,
          "author_name": "kamleshshimpi78",
          "author_url": "",
          "post_date": "01/26/2021 05:48:21",
          "content": "<p>thank you for the suggestion</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1170198,
      "author_name": "zzzhangzz",
      "author_url": "",
      "post_date": "01/26/2021 03:58:03",
      "content": "<p>Open the running notebook and click &lt; output &gt; to view the output data. There is a downward arrow on the right of download, or click download all on the left.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1170184": "Hi there, I need little bit of help, I am saving the trained weights using checkpoint callback ,how can I download and upload my trained weights ? I am using EfficientNetB0 with keras.",
    "1170197": "Hello!\n\nFirst of all, make sure that your `ModelCheckpoint` callback actually saves your weights by checking your `/kaggle/working` directory.\nThen you can hit **Save Version** (Save & Run All (Commit)) in your notebook. By going to the **Output** section in the Viewer and hitting **+ New Dataset** you can create a new private dataset containing all the outputs of your notebook. The dataset can then be imported to your inference notebook. Or simply download the results from the **Output** section to use them on your local machine.",
    "1170198": "Open the running notebook and click < output > to view the output data. There is a downward arrow on the right of download, or click download all on the left.",
    "1170295": "thank you for the suggestion"
  },
  "source": "meta"
}