{
  "id": 165400,
  "title": "Scoring Script",
  "url": "/competitions/landmark-retrieval-2020/discussion/165400",
  "author_name": "",
  "post_date": "2020-07-09T15:17:58.806787500Z",
  "votes": 13,
  "comment_count": 6,
  "views": 0,
  "content": "<p>Hi all!</p>\n<p>The code used to load your SavedModel, create the embeddings, and score your submission is provided <a href=\"https://www.kaggle.com/philculliton/landmark-retrieval-2020-shared-scoring-script\" target=\"_blank\">here</a>. Please let us know if you have any questions.</p>",
  "messages": [
    {
      "id": "921809",
      "postDate": "07/09/2020 15:17:58",
      "content": "<p>Hi all!</p>\n<p>The code used to load your SavedModel, create the embeddings, and score your submission is provided <a href=\"https://www.kaggle.com/philculliton/landmark-retrieval-2020-shared-scoring-script\" target=\"_blank\">here</a>. Please let us know if you have any questions.</p>",
      "rawMarkdown": "Hi all!\n\nThe code used to load your SavedModel, create the embeddings, and score your submission is provided [here](https://www.kaggle.com/philculliton/landmark-retrieval-2020-shared-scoring-script). Please let us know if you have any questions.",
      "votes": null
    },
    {
      "id": "922556",
      "postDate": "07/10/2020 07:44:46",
      "content": "<p>Thank you for sharing!</p>\n\n<p>Locally I can generate the embeddings from my model using the scoring.py script. But when I submit it through a kernel I get the error: \"Evaluation Exception: The Evaluation Kernel encountered an unexpected error.\".</p>\n\n<p>Is there any way to get more specific information on the error?</p>\n\n<p>I've tried different output embeddings (100,) and (2048,) but no luck. Also removing negative predictions didn't help.</p>\n\n<p>Edit: Found and fixed the error. More feedback on errors would still be nice though.</p>",
      "rawMarkdown": "Thank you for sharing!\n\nLocally I can generate the embeddings from my model using the scoring.py script. But when I submit it through a kernel I get the error: \"Evaluation Exception: The Evaluation Kernel encountered an unexpected error.\".\n\nIs there any way to get more specific information on the error?\n\nI've tried different output embeddings (100,) and (2048,) but no luck. Also removing negative predictions didn't help.\n\nEdit: Found and fixed the error. More feedback on errors would still be nice though.",
      "votes": null
    },
    {
      "id": "928624",
      "postDate": "07/14/2020 05:32:33",
      "content": "<p>Thanks for sharing it's the first time , seeing this submission procedure.</p>",
      "rawMarkdown": "Thanks for sharing it's the first time , seeing this submission procedure.",
      "votes": null
    },
    {
      "id": "952821",
      "postDate": "07/31/2020 08:40:58",
      "content": "<p><a href=\"/philculliton\">@philculliton</a> I use the code below to load the model and save it to another path. But I got nothing in my_model2/variables . Could you tell me how to save the model again?</p>\n\n<blockquote>\n  <p>model = tf.saved_model.load(str(saved_model_path))\n  tf.saved_model.save(model, export_dir=\"./my_model2\", signatures=model.signatures)</p>\n</blockquote>",
      "rawMarkdown": "philculliton I use the code below to load the model and save it to another path. But I got nothing in my_model2/variables . Could you tell me how to save the model again?\n\n&gt; model = tf.saved_model.load(str(saved_model_path))\ntf.saved_model.save(model, export_dir=\"./my_model2\", signatures=model.signatures)",
      "votes": null
    },
    {
      "id": "954691",
      "postDate": "08/02/2020 02:11:08",
      "content": "<p>Hi, I am not able infer the value of K here. You might have hidden it but i presume it is threshold. is it possible to know this value? does it exactly matches the index value of 100 to calculate the metric mAP@100?\nThanks</p>",
      "rawMarkdown": "Hi, I am not able infer the value of K here. You might have hidden it but i presume it is threshold. is it possible to know this value? does it exactly matches the index value of 100 to calculate the metric mAP@100?\nThanks",
      "votes": null
    },
    {
      "id": "960240",
      "postDate": "08/06/2020 08:55:57",
      "content": "<p>While scoring, will the images be passed as RGB or BGR??</p>",
      "rawMarkdown": "While scoring, will the images be passed as RGB or BGR??",
      "votes": null
    },
    {
      "id": "960753",
      "postDate": "08/06/2020 16:49:15",
      "content": "<p>Rgb</p>",
      "rawMarkdown": "Rgb",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 922556,
      "author_name": "reflexion",
      "author_url": "",
      "post_date": "07/10/2020 07:44:46",
      "content": "<p>Thank you for sharing!</p>\n\n<p>Locally I can generate the embeddings from my model using the scoring.py script. But when I submit it through a kernel I get the error: \"Evaluation Exception: The Evaluation Kernel encountered an unexpected error.\".</p>\n\n<p>Is there any way to get more specific information on the error?</p>\n\n<p>I've tried different output embeddings (100,) and (2048,) but no luck. Also removing negative predictions didn't help.</p>\n\n<p>Edit: Found and fixed the error. More feedback on errors would still be nice though.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 928624,
      "author_name": "yash612",
      "author_url": "",
      "post_date": "07/14/2020 05:32:33",
      "content": "<p>Thanks for sharing it's the first time , seeing this submission procedure.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 952821,
      "author_name": "wuliaokaola",
      "author_url": "",
      "post_date": "07/31/2020 08:40:58",
      "content": "<p><a href=\"/philculliton\">@philculliton</a> I use the code below to load the model and save it to another path. But I got nothing in my_model2/variables . Could you tell me how to save the model again?</p>\n\n<blockquote>\n  <p>model = tf.saved_model.load(str(saved_model_path))\n  tf.saved_model.save(model, export_dir=\"./my_model2\", signatures=model.signatures)</p>\n</blockquote>",
      "votes": null,
      "replies": []
    },
    {
      "id": 954691,
      "author_name": "udaygurugubelli",
      "author_url": "",
      "post_date": "08/02/2020 02:11:08",
      "content": "<p>Hi, I am not able infer the value of K here. You might have hidden it but i presume it is threshold. is it possible to know this value? does it exactly matches the index value of 100 to calculate the metric mAP@100?\nThanks</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 960240,
      "author_name": "photondodo",
      "author_url": "",
      "post_date": "08/06/2020 08:55:57",
      "content": "<p>While scoring, will the images be passed as RGB or BGR??</p>",
      "votes": null,
      "replies": [
        {
          "id": 960753,
          "author_name": "stenford23",
          "author_url": "",
          "post_date": "08/06/2020 16:49:15",
          "content": "<p>Rgb</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "921809": "Hi all!\n\nThe code used to load your SavedModel, create the embeddings, and score your submission is provided [here](https://www.kaggle.com/philculliton/landmark-retrieval-2020-shared-scoring-script). Please let us know if you have any questions.",
    "922556": "Thank you for sharing!\n\nLocally I can generate the embeddings from my model using the scoring.py script. But when I submit it through a kernel I get the error: \"Evaluation Exception: The Evaluation Kernel encountered an unexpected error.\".\n\nIs there any way to get more specific information on the error?\n\nI've tried different output embeddings (100,) and (2048,) but no luck. Also removing negative predictions didn't help.\n\nEdit: Found and fixed the error. More feedback on errors would still be nice though.",
    "928624": "Thanks for sharing it's the first time , seeing this submission procedure.",
    "952821": "philculliton I use the code below to load the model and save it to another path. But I got nothing in my_model2/variables . Could you tell me how to save the model again?\n\n&gt; model = tf.saved_model.load(str(saved_model_path))\ntf.saved_model.save(model, export_dir=\"./my_model2\", signatures=model.signatures)",
    "954691": "Hi, I am not able infer the value of K here. You might have hidden it but i presume it is threshold. is it possible to know this value? does it exactly matches the index value of 100 to calculate the metric mAP@100?\nThanks",
    "960240": "While scoring, will the images be passed as RGB or BGR??",
    "960753": "Rgb"
  },
  "source": "meta"
}