{
  "id": 409898,
  "title": "Python 3.10 Update Is Complete",
  "url": "/competitions/predict-student-performance-from-game-play/discussion/409898",
  "author_name": "Phil Culliton",
  "post_date": "2023-05-13T02:00:34.493000",
  "votes": 9,
  "comment_count": 25,
  "views": 0,
  "content": "<p>Hi all - we've completed rebuilding the competition's Timeseries API in Python 3.10. Instructions for using the new API are below.</p>\n<h2>Working with the API and Python 3.10</h2>\n<p>Kaggle notebooks now support Python 3.10. Given the way that this competition's API works, each binary is only compatible with a single version of Python so we've added a second copy of the API (<code>jo_wilder_310</code>). If you're using a new notebook, you just need to import the API using that name.</p>\n<p>Notebooks that worked under the old API shouldn't need to change at all.</p>",
  "messages": [
    {
      "id": 2257072,
      "postDate": "2023-05-13T02:00:34.493Z",
      "content": "<p>Hi all - we've completed rebuilding the competition's Timeseries API in Python 3.10. Instructions for using the new API are below.</p>\n<h2>Working with the API and Python 3.10</h2>\n<p>Kaggle notebooks now support Python 3.10. Given the way that this competition's API works, each binary is only compatible with a single version of Python so we've added a second copy of the API (<code>jo_wilder_310</code>). If you're using a new notebook, you just need to import the API using that name.</p>\n<p>Notebooks that worked under the old API shouldn't need to change at all.</p>",
      "rawMarkdown": "Hi all - we've completed rebuilding the competition's Timeseries API in Python 3.10. Instructions for using the new API are below.\n\n##Working with the API and Python 3.10\nKaggle notebooks now support Python 3.10. Given the way that this competition's API works, each binary is only compatible with a single version of Python so we've added a second copy of the API (`jo_wilder_310`). If you're using a new notebook, you just need to import the API using that name.\n\nNotebooks that worked under the old API shouldn't need to change at all.",
      "votes": 9
    },
    {
      "id": 2264279,
      "postDate": "2023-05-18T09:39:07.270Z",
      "content": "<p><a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> I want to know that why I used the jo_wilder_310 api to predict my previous code, the lb score is lower more than 0.04?</p>",
      "rawMarkdown": "@philculliton I want to know that why I used the jo_wilder_310 api to predict my previous code, the lb score is lower more than 0.04?",
      "votes": 5,
      "replies": [
        {
          "id": 2266542,
          "postDate": "2023-05-20T07:23:34.027Z",
          "content": "<p>I encountered the same phenomenon. My 0.703 submission resulted in a LB score of 0.548 with the jo_wilder_310 api.</p>\n<p>However, I found that I am partially to blame here. My original code assumed that the latter data passed from the API was already sorted by question number, which was true in original API, but <strong>in jo_wilder_310, sorting seemed to be not guaranteed</strong>, because after adding the sorting process, the LB score improved to 0.679.</p>\n<p>However, it still remains lower than the original result. While it is possible that this is not due to a bug in the API, I consider it highly doubtful. If anyone has been able to reproduce the results with the new API, I would appreciate it if you could let me know.</p>",
          "rawMarkdown": "I encountered the same phenomenon. My 0.703 submission resulted in a LB score of 0.548 with the jo_wilder_310 api.\n\nHowever, I found that I am partially to blame here. My original code assumed that the latter data passed from the API was already sorted by question number, which was true in original API, but **in jo_wilder_310, sorting seemed to be not guaranteed**, because after adding the sorting process, the LB score improved to 0.679.\n\nHowever, it still remains lower than the original result. While it is possible that this is not due to a bug in the API, I consider it highly doubtful. If anyone has been able to reproduce the results with the new API, I would appreciate it if you could let me know.",
          "votes": 11,
          "replies": [
            {
              "id": 2266604,
              "postDate": "2023-05-20T08:05:19.550Z",
              "content": "<p><a href=\"https://www.kaggle.com/rsakata\" target=\"_blank\">@rsakata</a> <br>\nThanks for sharing.  <br>\nIs this talking about the level_group order? Or are you talking about the sample_submission data output from the API not being sorted by question_no?</p>",
              "rawMarkdown": "@rsakata \nThanks for sharing.  \nIs this talking about the level_group order? Or are you talking about the sample_submission data output from the API not being sorted by question_no?"
            },
            {
              "id": 2266625,
              "postDate": "2023-05-20T08:23:54.203Z",
              "content": "<p>I think he's probably sorting with 'elapsed time'. … </p>",
              "rawMarkdown": "I think he's probably sorting with 'elapsed time'. ... "
            },
            {
              "id": 2266636,
              "postDate": "2023-05-20T08:31:52.357Z",
              "rawMarkdown": "",
              "isDeleted": true
            },
            {
              "id": 2266637,
              "postDate": "2023-05-20T08:32:10.220Z",
              "content": "<p><a href=\"https://www.kaggle.com/t88take\" target=\"_blank\">@t88take</a> <br>\nThe latter is true. I meant the question number that is concatenated with the session id.</p>",
              "rawMarkdown": "@t88take \nThe latter is true. I meant the question number that is concatenated with the session id.",
              "votes": 3
            },
            {
              "id": 2266921,
              "postDate": "2023-05-20T13:22:59.063Z",
              "content": "<p><a href=\"https://www.kaggle.com/rsakata\" target=\"_blank\">@rsakata</a> you are not to blame that the APIs are not the same. someone who did not test these APIs is</p>",
              "rawMarkdown": "@rsakata you are not to blame that the APIs are not the same. someone who did not test these APIs is"
            },
            {
              "id": 2268036,
              "postDate": "2023-05-21T12:09:58.450Z",
              "content": "<p><a href=\"https://www.kaggle.com/rsakata\" target=\"_blank\">@rsakata</a>, <a href=\"https://www.kaggle.com/juzqyxs\" target=\"_blank\">@juzqyxs</a>  , did you find why the score lowered ~0.04? I have the same situation. Should we stick with the old version of jo_wilder?</p>",
              "rawMarkdown": "@rsakata, @juzqyxs  , did you find why the score lowered ~0.04? I have the same situation. Should we stick with the old version of jo_wilder?",
              "votes": 1
            },
            {
              "id": 2268068,
              "postDate": "2023-05-21T12:39:19.320Z",
              "content": "<p>the py37 of jo_wilder is ok.</p>",
              "rawMarkdown": "the py37 of jo_wilder is ok.",
              "votes": 2
            },
            {
              "id": 2268335,
              "postDate": "2023-05-21T16:44:52.363Z",
              "content": "<p>Can you please guide how can i sort for jo_wilder_310 api ?</p>",
              "rawMarkdown": "Can you please guide how can i sort for jo_wilder_310 api ?\n"
            },
            {
              "id": 2269750,
              "postDate": "2023-05-22T17:17:18.617Z",
              "content": "<p>Exactly same, my 0.697 lb resulted in a 0.5 lb score. <a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> any updates regarding this?</p>",
              "rawMarkdown": "Exactly same, my 0.697 lb resulted in a 0.5 lb score. @philculliton any updates regarding this?",
              "votes": 1
            },
            {
              "id": 2270115,
              "postDate": "2023-05-23T02:50:23.763Z",
              "content": "<p>I also compared jo_wilder and jo_wilder310 with the same code and both were LB0.695.  <br>\nI suspected a problem with the order of the data output by iter, so I added the following check code and ran it, but no error occurred in either version.</p>\n<pre><code>prev_level_group = None\nprev_session_id = None\nfor (test_sessions, sample_submission) in iter_test:\n    level_group = test_sessions[\"level_group\"].values[0]\n    session_id = test_sessions[\"session_id\"].values[0]\n    if prev_level_group is None:\n        assert level_group == \"0-4\"\n    else:\n        if level_group == \"0-4\":\n            assert prev_level_group == \"13-22\"\n            assert session_id != prev_session_id\n        elif level_group == \"5-12\":\n            assert prev_level_group == \"0-4\"\n            assert session_id == prev_session_id\n        elif level_group == \"13-22\":\n            assert prev_level_group == \"5-12\"\n            assert session_id == prev_session_id\n    prev_level_group = level_group\n    prev_session_id = session_id\n</code></pre>",
              "rawMarkdown": "I also compared jo_wilder and jo_wilder310 with the same code and both were LB0.695.  \nI suspected a problem with the order of the data output by iter, so I added the following check code and ran it, but no error occurred in either version.\n\n    prev_level_group = None\n    prev_session_id = None\n    for (test_sessions, sample_submission) in iter_test:\n        level_group = test_sessions[\"level_group\"].values[0]\n        session_id = test_sessions[\"session_id\"].values[0]\n        if prev_level_group is None:\n            assert level_group == \"0-4\"\n        else:\n            if level_group == \"0-4\":\n                assert prev_level_group == \"13-22\"\n                assert session_id != prev_session_id\n            elif level_group == \"5-12\":\n                assert prev_level_group == \"0-4\"\n                assert session_id == prev_session_id\n            elif level_group == \"13-22\":\n                assert prev_level_group == \"5-12\"\n                assert session_id == prev_session_id\n        prev_level_group = level_group\n        prev_session_id = session_id",
              "votes": 3
            },
            {
              "id": 2271587,
              "postDate": "2023-05-24T03:21:00.147Z",
              "content": "<p>Thanks everyone for the info. I posted my conclusion on this issue in a new discussion thread (<a href=\"https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/412512\" target=\"_blank\">here</a>) as I thought it should be widely known to the participants.</p>",
              "rawMarkdown": "Thanks everyone for the info. I posted my conclusion on this issue in a new discussion thread ([here](https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/412512)) as I thought it should be widely known to the participants.",
              "votes": 3
            }
          ]
        }
      ]
    },
    {
      "id": 2284723,
      "postDate": "2023-06-02T08:37:25.160Z",
      "content": "<p>I submitted my notebook, which scored .7xx, a week ago. However, the fork of the notebook scored .63x (I am using the old competition API). <br>\nHas there been any change in the competition API?</p>",
      "rawMarkdown": "I submitted my notebook, which scored .7xx, a week ago. However, the fork of the notebook scored .63x (I am using the old competition API). \nHas there been any change in the competition API?",
      "votes": 1,
      "replies": [
        {
          "id": 2284812,
          "postDate": "2023-06-02T09:35:46.380Z",
          "content": "<p><a href=\"https://www.kaggle.com/chaudharypriyanshu\" target=\"_blank\">@chaudharypriyanshu</a> ,  I fixed setting <code>test = test.sort_values(by=\"index\").reset_index(drop=True)</code></p>\n<p>You should check <a href=\"https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/414109\" target=\"_blank\">https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/414109</a></p>",
          "rawMarkdown": "@chaudharypriyanshu ,  I fixed setting `test = test.sort_values(by=\"index\").reset_index(drop=True)`\n\nYou should check https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/414109",
          "votes": 1,
          "replies": [
            {
              "id": 2284834,
              "postDate": "2023-06-02T09:52:32.927Z",
              "content": "<p>Hi <a href=\"https://www.kaggle.com/steubk\" target=\"_blank\">@steubk</a> , Thanks for letting me know.</p>",
              "rawMarkdown": "Hi @steubk , Thanks for letting me know.",
              "votes": 1
            }
          ]
        }
      ]
    },
    {
      "id": 2264292,
      "postDate": "2023-05-18T09:50:26.820Z",
      "content": "<p>Is it mandatory to change the environment to python 3.10? Or can we still use python 3.7 for final submission?</p>",
      "rawMarkdown": "Is it mandatory to change the environment to python 3.10? Or can we still use python 3.7 for final submission?",
      "replies": [
        {
          "id": 2268018,
          "postDate": "2023-05-21T11:49:38.817Z",
          "content": "<p>hope this could help:</p>\n<blockquote>\n  <p>Hi all - we've completed rebuilding the competition's Timeseries API in Python 3.10. Instructions for using the new API are below.</p>\n  <h2>Working with the API and Python 3.10</h2>\n  <p>Kaggle notebooks now support Python 3.10. Given the way that this competition's API works, each binary is only compatible with a single version of Python so we've added a second copy of the API (<code>jo_wilder_310</code>). If you're using a new notebook, you just need to import the API using that name.</p>\n  <p>==&gt;<strong>Notebooks that worked under the old API shouldn't need to change at all.</strong></p>\n</blockquote>",
          "rawMarkdown": "hope this could help:\n\n> Hi all - we've completed rebuilding the competition's Timeseries API in Python 3.10. Instructions for using the new API are below.\n> \n> ##Working with the API and Python 3.10\n> Kaggle notebooks now support Python 3.10. Given the way that this competition's API works, each binary is only compatible with a single version of Python so we've added a second copy of the API (`jo_wilder_310`). If you're using a new notebook, you just need to import the API using that name.\n> \n> ==>**Notebooks that worked under the old API shouldn't need to change at all.**\n\n",
          "votes": 1,
          "replies": [
            {
              "id": 2268707,
              "postDate": "2023-05-21T23:57:55.253Z",
              "content": "<p>Thank you. Just wanted to make sure about it. </p>",
              "rawMarkdown": "Thank you. Just wanted to make sure about it. ",
              "votes": 1
            }
          ]
        }
      ]
    },
    {
      "id": 2257558,
      "postDate": "2023-05-13T13:19:59.653Z",
      "content": "<p>Perfect, thanks! This competition is an awesome pratice opportunity.</p>",
      "rawMarkdown": "Perfect, thanks! This competition is an awesome pratice opportunity."
    },
    {
      "id": 2257362,
      "postDate": "2023-05-13T09:41:22.303Z",
      "content": "<p>Thank you for sharing! This is interesting! </p>",
      "rawMarkdown": "Thank you for sharing! This is interesting! "
    },
    {
      "id": 2309604,
      "postDate": "2023-06-19T18:23:12.010Z",
      "content": "<p>Thanks for Sharing That Sir</p>",
      "rawMarkdown": "Thanks for Sharing That Sir"
    },
    {
      "id": 2281575,
      "postDate": "2023-05-31T01:55:19.203Z",
      "content": "<p>first comit thanks</p>",
      "rawMarkdown": "first comit thanks"
    },
    {
      "id": 2270994,
      "postDate": "2023-05-23T15:03:52.807Z",
      "content": "<p>Thanks for sharing!!!</p>",
      "rawMarkdown": "Thanks for sharing!!!"
    },
    {
      "id": 2271132,
      "postDate": "2023-05-23T16:34:54.390Z",
      "content": "<p>first comit thanks</p>",
      "rawMarkdown": "first comit thanks",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 2264279,
      "author_name": "qyxs",
      "author_url": "",
      "post_date": "2023-05-18T09:39:07.270000",
      "content": "<p><a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> I want to know that why I used the jo_wilder_310 api to predict my previous code, the lb score is lower more than 0.04?</p>",
      "votes": 5,
      "replies": [
        {
          "id": 2266542,
          "author_name": "Jack (Japan)",
          "author_url": "",
          "post_date": "2023-05-20T07:23:34.027000",
          "content": "<p>I encountered the same phenomenon. My 0.703 submission resulted in a LB score of 0.548 with the jo_wilder_310 api.</p>\n<p>However, I found that I am partially to blame here. My original code assumed that the latter data passed from the API was already sorted by question number, which was true in original API, but <strong>in jo_wilder_310, sorting seemed to be not guaranteed</strong>, because after adding the sorting process, the LB score improved to 0.679.</p>\n<p>However, it still remains lower than the original result. While it is possible that this is not due to a bug in the API, I consider it highly doubtful. If anyone has been able to reproduce the results with the new API, I would appreciate it if you could let me know.</p>",
          "votes": 11,
          "replies": [
            {
              "id": 2266604,
              "author_name": "T88",
              "author_url": "",
              "post_date": "2023-05-20T08:05:19.550000",
              "content": "<p><a href=\"https://www.kaggle.com/rsakata\" target=\"_blank\">@rsakata</a> <br>\nThanks for sharing.  <br>\nIs this talking about the level_group order? Or are you talking about the sample_submission data output from the API not being sorted by question_no?</p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2266625,
              "author_name": "IsaacSim",
              "author_url": "",
              "post_date": "2023-05-20T08:23:54.203000",
              "content": "<p>I think he's probably sorting with 'elapsed time'. … </p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2266636,
              "author_name": "",
              "author_url": "",
              "post_date": "2023-05-20T08:31:52.357000",
              "content": "",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2266637,
              "author_name": "Jack (Japan)",
              "author_url": "",
              "post_date": "2023-05-20T08:32:10.220000",
              "content": "<p><a href=\"https://www.kaggle.com/t88take\" target=\"_blank\">@t88take</a> <br>\nThe latter is true. I meant the question number that is concatenated with the session id.</p>",
              "votes": 3,
              "replies": []
            },
            {
              "id": 2266921,
              "author_name": "empty",
              "author_url": "",
              "post_date": "2023-05-20T13:22:59.063000",
              "content": "<p><a href=\"https://www.kaggle.com/rsakata\" target=\"_blank\">@rsakata</a> you are not to blame that the APIs are not the same. someone who did not test these APIs is</p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2268036,
              "author_name": "Elias",
              "author_url": "",
              "post_date": "2023-05-21T12:09:58.450000",
              "content": "<p><a href=\"https://www.kaggle.com/rsakata\" target=\"_blank\">@rsakata</a>, <a href=\"https://www.kaggle.com/juzqyxs\" target=\"_blank\">@juzqyxs</a>  , did you find why the score lowered ~0.04? I have the same situation. Should we stick with the old version of jo_wilder?</p>",
              "votes": 1,
              "replies": []
            },
            {
              "id": 2268068,
              "author_name": "qyxs",
              "author_url": "",
              "post_date": "2023-05-21T12:39:19.320000",
              "content": "<p>the py37 of jo_wilder is ok.</p>",
              "votes": 2,
              "replies": []
            },
            {
              "id": 2268335,
              "author_name": "Aman Kankriya",
              "author_url": "",
              "post_date": "2023-05-21T16:44:52.363000",
              "content": "<p>Can you please guide how can i sort for jo_wilder_310 api ?</p>",
              "votes": 0,
              "replies": []
            },
            {
              "id": 2269750,
              "author_name": "pjmathematician",
              "author_url": "",
              "post_date": "2023-05-22T17:17:18.617000",
              "content": "<p>Exactly same, my 0.697 lb resulted in a 0.5 lb score. <a href=\"https://www.kaggle.com/philculliton\" target=\"_blank\">@philculliton</a> any updates regarding this?</p>",
              "votes": 1,
              "replies": []
            },
            {
              "id": 2270115,
              "author_name": "T88",
              "author_url": "",
              "post_date": "2023-05-23T02:50:23.763000",
              "content": "<p>I also compared jo_wilder and jo_wilder310 with the same code and both were LB0.695.  <br>\nI suspected a problem with the order of the data output by iter, so I added the following check code and ran it, but no error occurred in either version.</p>\n<pre><code>prev_level_group = None\nprev_session_id = None\nfor (test_sessions, sample_submission) in iter_test:\n    level_group = test_sessions[\"level_group\"].values[0]\n    session_id = test_sessions[\"session_id\"].values[0]\n    if prev_level_group is None:\n        assert level_group == \"0-4\"\n    else:\n        if level_group == \"0-4\":\n            assert prev_level_group == \"13-22\"\n            assert session_id != prev_session_id\n        elif level_group == \"5-12\":\n            assert prev_level_group == \"0-4\"\n            assert session_id == prev_session_id\n        elif level_group == \"13-22\":\n            assert prev_level_group == \"5-12\"\n            assert session_id == prev_session_id\n    prev_level_group = level_group\n    prev_session_id = session_id\n</code></pre>",
              "votes": 3,
              "replies": []
            },
            {
              "id": 2271587,
              "author_name": "Jack (Japan)",
              "author_url": "",
              "post_date": "2023-05-24T03:21:00.147000",
              "content": "<p>Thanks everyone for the info. I posted my conclusion on this issue in a new discussion thread (<a href=\"https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/412512\" target=\"_blank\">here</a>) as I thought it should be widely known to the participants.</p>",
              "votes": 3,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 2284723,
      "author_name": "Priyanshu Chaudhary",
      "author_url": "",
      "post_date": "2023-06-02T08:37:25.160000",
      "content": "<p>I submitted my notebook, which scored .7xx, a week ago. However, the fork of the notebook scored .63x (I am using the old competition API). <br>\nHas there been any change in the competition API?</p>",
      "votes": 1,
      "replies": [
        {
          "id": 2284812,
          "author_name": "steubk",
          "author_url": "",
          "post_date": "2023-06-02T09:35:46.380000",
          "content": "<p><a href=\"https://www.kaggle.com/chaudharypriyanshu\" target=\"_blank\">@chaudharypriyanshu</a> ,  I fixed setting <code>test = test.sort_values(by=\"index\").reset_index(drop=True)</code></p>\n<p>You should check <a href=\"https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/414109\" target=\"_blank\">https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/414109</a></p>",
          "votes": 1,
          "replies": [
            {
              "id": 2284834,
              "author_name": "Priyanshu Chaudhary",
              "author_url": "",
              "post_date": "2023-06-02T09:52:32.927000",
              "content": "<p>Hi <a href=\"https://www.kaggle.com/steubk\" target=\"_blank\">@steubk</a> , Thanks for letting me know.</p>",
              "votes": 1,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 2264292,
      "author_name": "IsaacSim",
      "author_url": "",
      "post_date": "2023-05-18T09:50:26.820000",
      "content": "<p>Is it mandatory to change the environment to python 3.10? Or can we still use python 3.7 for final submission?</p>",
      "votes": 0,
      "replies": [
        {
          "id": 2268018,
          "author_name": "steubk",
          "author_url": "",
          "post_date": "2023-05-21T11:49:38.817000",
          "content": "<p>hope this could help:</p>\n<blockquote>\n  <p>Hi all - we've completed rebuilding the competition's Timeseries API in Python 3.10. Instructions for using the new API are below.</p>\n  <h2>Working with the API and Python 3.10</h2>\n  <p>Kaggle notebooks now support Python 3.10. Given the way that this competition's API works, each binary is only compatible with a single version of Python so we've added a second copy of the API (<code>jo_wilder_310</code>). If you're using a new notebook, you just need to import the API using that name.</p>\n  <p>==&gt;<strong>Notebooks that worked under the old API shouldn't need to change at all.</strong></p>\n</blockquote>",
          "votes": 1,
          "replies": [
            {
              "id": 2268707,
              "author_name": "IsaacSim",
              "author_url": "",
              "post_date": "2023-05-21T23:57:55.253000",
              "content": "<p>Thank you. Just wanted to make sure about it. </p>",
              "votes": 1,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 2257558,
      "author_name": "Rodrigo Ramiro",
      "author_url": "",
      "post_date": "2023-05-13T13:19:59.653000",
      "content": "<p>Perfect, thanks! This competition is an awesome pratice opportunity.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2257362,
      "author_name": "Ericka42",
      "author_url": "",
      "post_date": "2023-05-13T09:41:22.303000",
      "content": "<p>Thank you for sharing! This is interesting! </p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2309604,
      "author_name": "Saha Chan",
      "author_url": "",
      "post_date": "2023-06-19T18:23:12.010000",
      "content": "<p>Thanks for Sharing That Sir</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2281575,
      "author_name": "Difu Tan",
      "author_url": "",
      "post_date": "2023-05-31T01:55:19.203000",
      "content": "<p>first comit thanks</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2270994,
      "author_name": "Jingru Su",
      "author_url": "",
      "post_date": "2023-05-23T15:03:52.807000",
      "content": "<p>Thanks for sharing!!!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2271132,
      "author_name": "",
      "author_url": "",
      "post_date": "2023-05-23T16:34:54.390000",
      "content": "<p>first comit thanks</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2257072": "Hi all - we've completed rebuilding the competition's Timeseries API in Python 3.10. Instructions for using the new API are below.\n\n##Working with the API and Python 3.10\nKaggle notebooks now support Python 3.10. Given the way that this competition's API works, each binary is only compatible with a single version of Python so we've added a second copy of the API (`jo_wilder_310`). If you're using a new notebook, you just need to import the API using that name.\n\nNotebooks that worked under the old API shouldn't need to change at all.",
    "2264279": "@philculliton I want to know that why I used the jo_wilder_310 api to predict my previous code, the lb score is lower more than 0.04?",
    "2284723": "I submitted my notebook, which scored .7xx, a week ago. However, the fork of the notebook scored .63x (I am using the old competition API). \nHas there been any change in the competition API?",
    "2264292": "Is it mandatory to change the environment to python 3.10? Or can we still use python 3.7 for final submission?",
    "2257558": "Perfect, thanks! This competition is an awesome pratice opportunity.",
    "2257362": "Thank you for sharing! This is interesting! ",
    "2309604": "Thanks for Sharing That Sir",
    "2281575": "first comit thanks",
    "2270994": "Thanks for sharing!!!",
    "2271132": "first comit thanks"
  }
}