{
  "id": 82755,
  "title": "sampleSubmission.csv",
  "url": "/competitions/cifar-10/discussion/82755",
  "author_name": "",
  "post_date": "2019-03-04T02:45:22.401891100Z",
  "votes": null,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Does anyone able to provide sampleSubmission.csv for test_images ? As the upload sampleSubmission.csv is all with label: cat for 300k images.</p>",
  "messages": [
    {
      "id": "483035",
      "postDate": "03/04/2019 02:45:22",
      "content": "<p>Does anyone able to provide sampleSubmission.csv for test_images ? As the upload sampleSubmission.csv is all with label: cat for 300k images.</p>",
      "rawMarkdown": "Does anyone able to provide sampleSubmission.csv for test_images ? As the upload sampleSubmission.csv is all with label: cat for 300k images.",
      "votes": null
    },
    {
      "id": "635732",
      "postDate": "09/28/2019 05:44:51",
      "content": "<p>All you need is to predict the 10 classes through your model.\nAnd the prediction answers on the sumbission.csv can be any word under below:\n- airplane\n- automobile \n- bird\n- cat \n- deer \n- dog \n- frog \n- horse \n- ship \n- truck</p>",
      "rawMarkdown": "All you need is to predict the 10 classes through your model.\nAnd the prediction answers on the sumbission.csv can be any word under below:\n- airplane\n- automobile \n- bird\n- cat \n- deer \n- dog \n- frog \n- horse \n- ship \n- truck",
      "votes": null
    },
    {
      "id": "777546",
      "postDate": "03/17/2020 19:01:32",
      "content": "<p>Hi there all, can someone explain how to create the submission file to upload, I am new to this, thanks.</p>",
      "rawMarkdown": "Hi there all, can someone explain how to create the submission file to upload, I am new to this, thanks.",
      "votes": null
    },
    {
      "id": "779550",
      "postDate": "03/19/2020 13:19:53",
      "content": "<p>Basically, your predict result would be the number 0~9 (10 classes).\nTherefore, to produce the submission file you have to convert the number to the relative words.</p>",
      "rawMarkdown": "Basically, your predict result would be the number 0~9 (10 classes).\nTherefore, to produce the submission file you have to convert the number to the relative words.",
      "votes": null
    },
    {
      "id": "780951",
      "postDate": "03/20/2020 19:12:03",
      "content": "<p>Hi <a href=\"/polunwang\">@polunwang</a>, thanks for the replay I am using this code:\n```</p>\n\n<h1>loop over the model paths, loading the model, and adding it to</h1>\n\n<h1>the list of models</h1>\n\n<p>for (i, modelPath) in enumerate(modelPaths):\n    print(\"[INFO] loading model {}/{}\".format(i + 1,len(modelPaths)))\n    models.append(load_model(modelPath))</p>\n\n<h1>initialize the list of predictions</h1>\n\n<p>print(\"[INFO] evaluating ensemble...\")\npredictions = []</p>\n\n<h1>loop over the models</h1>\n\n<p>for model in models:\n    # use the current model to make predictions on the testing data,\n    # then store these predictions in the aggregate predictions list\n    predictions.append(model.predict(testX, batch_size=64))</p>\n\n<h1>average the probabilities across all model predictions, then show</h1>\n\n<h1>a classification report</h1>\n\n<p>predictions = np.average(predictions, axis=0)\n```</p>\n\n<p>Sorry to ask you, but what do I do next?</p>",
      "rawMarkdown": "Hi @polunwang, thanks for the replay I am using this code:\n```\n# loop over the model paths, loading the model, and adding it to\n# the list of models\nfor (i, modelPath) in enumerate(modelPaths):\n\tprint(\"[INFO] loading model {}/{}\".format(i + 1,len(modelPaths)))\n\tmodels.append(load_model(modelPath))\n\n# initialize the list of predictions\nprint(\"[INFO] evaluating ensemble...\")\npredictions = []\n\n# loop over the models\nfor model in models:\n\t# use the current model to make predictions on the testing data,\n\t# then store these predictions in the aggregate predictions list\n\tpredictions.append(model.predict(testX, batch_size=64))\n\n# average the probabilities across all model predictions, then show\n# a classification report\npredictions = np.average(predictions, axis=0)\n```\n\nSorry to ask you, but what do I do next?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 635732,
      "author_name": "polunwang",
      "author_url": "",
      "post_date": "09/28/2019 05:44:51",
      "content": "<p>All you need is to predict the 10 classes through your model.\nAnd the prediction answers on the sumbission.csv can be any word under below:\n- airplane\n- automobile \n- bird\n- cat \n- deer \n- dog \n- frog \n- horse \n- ship \n- truck</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 777546,
      "author_name": "oscarrangel",
      "author_url": "",
      "post_date": "03/17/2020 19:01:32",
      "content": "<p>Hi there all, can someone explain how to create the submission file to upload, I am new to this, thanks.</p>",
      "votes": null,
      "replies": [
        {
          "id": 779550,
          "author_name": "polunwang",
          "author_url": "",
          "post_date": "03/19/2020 13:19:53",
          "content": "<p>Basically, your predict result would be the number 0~9 (10 classes).\nTherefore, to produce the submission file you have to convert the number to the relative words.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 780951,
          "author_name": "oscarrangel",
          "author_url": "",
          "post_date": "03/20/2020 19:12:03",
          "content": "<p>Hi <a href=\"/polunwang\">@polunwang</a>, thanks for the replay I am using this code:\n```</p>\n\n<h1>loop over the model paths, loading the model, and adding it to</h1>\n\n<h1>the list of models</h1>\n\n<p>for (i, modelPath) in enumerate(modelPaths):\n    print(\"[INFO] loading model {}/{}\".format(i + 1,len(modelPaths)))\n    models.append(load_model(modelPath))</p>\n\n<h1>initialize the list of predictions</h1>\n\n<p>print(\"[INFO] evaluating ensemble...\")\npredictions = []</p>\n\n<h1>loop over the models</h1>\n\n<p>for model in models:\n    # use the current model to make predictions on the testing data,\n    # then store these predictions in the aggregate predictions list\n    predictions.append(model.predict(testX, batch_size=64))</p>\n\n<h1>average the probabilities across all model predictions, then show</h1>\n\n<h1>a classification report</h1>\n\n<p>predictions = np.average(predictions, axis=0)\n```</p>\n\n<p>Sorry to ask you, but what do I do next?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "483035": "Does anyone able to provide sampleSubmission.csv for test_images ? As the upload sampleSubmission.csv is all with label: cat for 300k images.",
    "635732": "All you need is to predict the 10 classes through your model.\nAnd the prediction answers on the sumbission.csv can be any word under below:\n- airplane\n- automobile \n- bird\n- cat \n- deer \n- dog \n- frog \n- horse \n- ship \n- truck",
    "777546": "Hi there all, can someone explain how to create the submission file to upload, I am new to this, thanks.",
    "779550": "Basically, your predict result would be the number 0~9 (10 classes).\nTherefore, to produce the submission file you have to convert the number to the relative words.",
    "780951": "Hi @polunwang, thanks for the replay I am using this code:\n```\n# loop over the model paths, loading the model, and adding it to\n# the list of models\nfor (i, modelPath) in enumerate(modelPaths):\n\tprint(\"[INFO] loading model {}/{}\".format(i + 1,len(modelPaths)))\n\tmodels.append(load_model(modelPath))\n\n# initialize the list of predictions\nprint(\"[INFO] evaluating ensemble...\")\npredictions = []\n\n# loop over the models\nfor model in models:\n\t# use the current model to make predictions on the testing data,\n\t# then store these predictions in the aggregate predictions list\n\tpredictions.append(model.predict(testX, batch_size=64))\n\n# average the probabilities across all model predictions, then show\n# a classification report\npredictions = np.average(predictions, axis=0)\n```\n\nSorry to ask you, but what do I do next?"
  },
  "source": "meta"
}