{
  "id": 106579,
  "title": "Unknown initializer: EfficientConv2DKernelInitializer",
  "url": "/competitions/aptos2019-blindness-detection/discussion/106579",
  "author_name": "",
  "post_date": "2019-08-30T04:00:20.563977500Z",
  "votes": null,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I am getting Unknown initializer: EfficientConv2DKernelInitializer error while loading my model in inference script.\nCan anybody have any suggestions to resolve it...\nplease help me...</p>",
  "messages": [
    {
      "id": "612878",
      "postDate": "08/30/2019 04:00:20",
      "content": "<p>I am getting Unknown initializer: EfficientConv2DKernelInitializer error while loading my model in inference script.\nCan anybody have any suggestions to resolve it...\nplease help me...</p>",
      "rawMarkdown": "I am getting Unknown initializer: EfficientConv2DKernelInitializer error while loading my model in inference script.\nCan anybody have any suggestions to resolve it...\nplease help me...",
      "votes": null
    },
    {
      "id": "614214",
      "postDate": "08/31/2019 08:10:20",
      "content": "<p>You are probably trying to load your model like this:\n<code>from keras.models import load_model</code></p>\n\n<p>but should be:\n<code>from efficientnet import load_model</code></p>",
      "rawMarkdown": "You are probably trying to load your model like this:\n`from keras.models import load_model`\n\n but should be:\n` from efficientnet import load_model`",
      "votes": null
    },
    {
      "id": "614327",
      "postDate": "08/31/2019 11:08:10",
      "content": "<p>Internet is needed to install efficientnet but, I can't ON internet as this script is going to generate my submission file.</p>",
      "rawMarkdown": "Internet is needed to install efficientnet but, I can't ON internet as this script is going to generate my submission file.",
      "votes": null
    },
    {
      "id": "614629",
      "postDate": "08/31/2019 19:50:34",
      "content": "<p>Got it!\n1. Use this way to test your submission, generate your CSV locally and then upload with dataset on Kaggle, replacing in this Kernel way <a href=\"https://www.kaggle.com/danilds/fast-submission\">https://www.kaggle.com/danilds/fast-submission</a></p>\n\n<ol>\n<li>You can clone the repository, follow this kernel to learn how to do this <a href=\"https://www.kaggle.com/xhlulu/easily-load-efficientnet-keras\">https://www.kaggle.com/xhlulu/easily-load-efficientnet-keras</a></li>\n</ol>",
      "rawMarkdown": "Got it!\n1. Use this way to test your submission, generate your CSV locally and then upload with dataset on Kaggle, replacing in this Kernel way https://www.kaggle.com/danilds/fast-submission\n\n2. You can clone the repository, follow this kernel to learn how to do this https://www.kaggle.com/xhlulu/easily-load-efficientnet-keras",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 614214,
      "author_name": "custodiogabriel",
      "author_url": "",
      "post_date": "08/31/2019 08:10:20",
      "content": "<p>You are probably trying to load your model like this:\n<code>from keras.models import load_model</code></p>\n\n<p>but should be:\n<code>from efficientnet import load_model</code></p>",
      "votes": null,
      "replies": [
        {
          "id": 614327,
          "author_name": "niteshfre",
          "author_url": "",
          "post_date": "08/31/2019 11:08:10",
          "content": "<p>Internet is needed to install efficientnet but, I can't ON internet as this script is going to generate my submission file.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 614629,
          "author_name": "custodiogabriel",
          "author_url": "",
          "post_date": "08/31/2019 19:50:34",
          "content": "<p>Got it!\n1. Use this way to test your submission, generate your CSV locally and then upload with dataset on Kaggle, replacing in this Kernel way <a href=\"https://www.kaggle.com/danilds/fast-submission\">https://www.kaggle.com/danilds/fast-submission</a></p>\n\n<ol>\n<li>You can clone the repository, follow this kernel to learn how to do this <a href=\"https://www.kaggle.com/xhlulu/easily-load-efficientnet-keras\">https://www.kaggle.com/xhlulu/easily-load-efficientnet-keras</a></li>\n</ol>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "612878": "I am getting Unknown initializer: EfficientConv2DKernelInitializer error while loading my model in inference script.\nCan anybody have any suggestions to resolve it...\nplease help me...",
    "614214": "You are probably trying to load your model like this:\n`from keras.models import load_model`\n\n but should be:\n` from efficientnet import load_model`",
    "614327": "Internet is needed to install efficientnet but, I can't ON internet as this script is going to generate my submission file.",
    "614629": "Got it!\n1. Use this way to test your submission, generate your CSV locally and then upload with dataset on Kaggle, replacing in this Kernel way https://www.kaggle.com/danilds/fast-submission\n\n2. You can clone the repository, follow this kernel to learn how to do this https://www.kaggle.com/xhlulu/easily-load-efficientnet-keras"
  },
  "source": "meta"
}