{
  "id": 214377,
  "title": "Model is not training need help!",
  "url": "/competitions/cassava-leaf-disease-classification/discussion/214377",
  "author_name": "",
  "post_date": "2021-01-26T12:20:04.394826700Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I am using EfficientNetB0 with keras, we cant keep internet on while submitting result, so I trained the model first using imagenet weights with keeping the internet on, then I save the weights using model checkpoint and then again trained the model using this saved own trained weights then I disable the internet for saving the version but I got the error ''URL fetch failure on <a href=\"https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5:\" target=\"_blank\">https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5:</a> None -- [Errno -3] Temporary failure in name resolution'' <br>\nhere is my model link-<a href=\"url\" target=\"_blank\">https://www.kaggle.com/kamleshshimpi78/leaf-disease-identification-keras</a><br>\nplease help with the issue.</p>",
  "messages": [
    {
      "id": "1170760",
      "postDate": "01/26/2021 12:20:04",
      "content": "<p>I am using EfficientNetB0 with keras, we cant keep internet on while submitting result, so I trained the model first using imagenet weights with keeping the internet on, then I save the weights using model checkpoint and then again trained the model using this saved own trained weights then I disable the internet for saving the version but I got the error ''URL fetch failure on <a href=\"https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5:\" target=\"_blank\">https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5:</a> None -- [Errno -3] Temporary failure in name resolution'' <br>\nhere is my model link-<a href=\"url\" target=\"_blank\">https://www.kaggle.com/kamleshshimpi78/leaf-disease-identification-keras</a><br>\nplease help with the issue.</p>",
      "rawMarkdown": "I am using EfficientNetB0 with keras, we cant keep internet on while submitting result, so I trained the model first using imagenet weights with keeping the internet on, then I save the weights using model checkpoint and then again trained the model using this saved own trained weights then I disable the internet for saving the version but I got the error ''URL fetch failure on https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5: None -- [Errno -3] Temporary failure in name resolution'' \nhere is my model link-[https://www.kaggle.com/kamleshshimpi78/leaf-disease-identification-keras](url)\nplease help with the issue.",
      "votes": null
    },
    {
      "id": "1170922",
      "postDate": "01/26/2021 14:30:23",
      "content": "<p>Hello!</p>\n<p>Just import <code>efficientnet</code> library source code to your notebook rather than trying to <code>pip install</code> it. There are a bunch of them uploaded on Kaggle. For instance, if you import <strong><a href=\"https://www.kaggle.com/awsaf49/efficientnet-keras-dataset\" target=\"_blank\">this one</a></strong>, add these lines to your notebook:<br>\n<code>import sys</code><br>\n<code>sys.path.append('/kaggle/input/efficientnet-keras-dataset/efficientnet_kaggle')</code><br>\nand then:<br>\n<code>import efficientnet.keras as efn</code><br>\njust as usual.</p>",
      "rawMarkdown": "Hello!\n\nJust import `efficientnet` library source code to your notebook rather than trying to `pip install` it. There are a bunch of them uploaded on Kaggle. For instance, if you import **[this one](https://www.kaggle.com/awsaf49/efficientnet-keras-dataset)**, add these lines to your notebook:\n`import sys`\n`sys.path.append('/kaggle/input/efficientnet-keras-dataset/efficientnet_kaggle')`\nand then:\n`import efficientnet.keras as efn`\njust as usual.",
      "votes": null
    },
    {
      "id": "1171042",
      "postDate": "01/26/2021 15:43:40",
      "content": "<p>Hello,<br>\n    To overcome this problem use 2 notebooks, one for training your model and then save the model. Add the model to a dataset.<br>\nanother notebook from which to predict. Keep internet off for the second one. Add the dataset to the second notebook that you made in the first step.</p>",
      "rawMarkdown": "Hello,\n    To overcome this problem use 2 notebooks, one for training your model and then save the model. Add the model to a dataset.\nanother notebook from which to predict. Keep internet off for the second one. Add the dataset to the second notebook that you made in the first step.",
      "votes": null
    },
    {
      "id": "1171823",
      "postDate": "01/27/2021 05:41:18",
      "content": "<p>thanks for help got the solution.</p>",
      "rawMarkdown": "thanks for help got the solution.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1170922,
      "author_name": "nickuzmenkov",
      "author_url": "",
      "post_date": "01/26/2021 14:30:23",
      "content": "<p>Hello!</p>\n<p>Just import <code>efficientnet</code> library source code to your notebook rather than trying to <code>pip install</code> it. There are a bunch of them uploaded on Kaggle. For instance, if you import <strong><a href=\"https://www.kaggle.com/awsaf49/efficientnet-keras-dataset\" target=\"_blank\">this one</a></strong>, add these lines to your notebook:<br>\n<code>import sys</code><br>\n<code>sys.path.append('/kaggle/input/efficientnet-keras-dataset/efficientnet_kaggle')</code><br>\nand then:<br>\n<code>import efficientnet.keras as efn</code><br>\njust as usual.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1171042,
      "author_name": "mohneesh7",
      "author_url": "",
      "post_date": "01/26/2021 15:43:40",
      "content": "<p>Hello,<br>\n    To overcome this problem use 2 notebooks, one for training your model and then save the model. Add the model to a dataset.<br>\nanother notebook from which to predict. Keep internet off for the second one. Add the dataset to the second notebook that you made in the first step.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1171823,
          "author_name": "kamleshshimpi78",
          "author_url": "",
          "post_date": "01/27/2021 05:41:18",
          "content": "<p>thanks for help got the solution.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1170760": "I am using EfficientNetB0 with keras, we cant keep internet on while submitting result, so I trained the model first using imagenet weights with keeping the internet on, then I save the weights using model checkpoint and then again trained the model using this saved own trained weights then I disable the internet for saving the version but I got the error ''URL fetch failure on https://storage.googleapis.com/keras-applications/efficientnetb0_notop.h5: None -- [Errno -3] Temporary failure in name resolution'' \nhere is my model link-[https://www.kaggle.com/kamleshshimpi78/leaf-disease-identification-keras](url)\nplease help with the issue.",
    "1170922": "Hello!\n\nJust import `efficientnet` library source code to your notebook rather than trying to `pip install` it. There are a bunch of them uploaded on Kaggle. For instance, if you import **[this one](https://www.kaggle.com/awsaf49/efficientnet-keras-dataset)**, add these lines to your notebook:\n`import sys`\n`sys.path.append('/kaggle/input/efficientnet-keras-dataset/efficientnet_kaggle')`\nand then:\n`import efficientnet.keras as efn`\njust as usual.",
    "1171042": "Hello,\n    To overcome this problem use 2 notebooks, one for training your model and then save the model. Add the model to a dataset.\nanother notebook from which to predict. Keep internet off for the second one. Add the dataset to the second notebook that you made in the first step.",
    "1171823": "thanks for help got the solution."
  },
  "source": "meta"
}