{
  "id": 94193,
  "title": "Does anyone know how to prevent catboost generating many irrelevant output files?",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/94193",
  "author_name": "",
  "post_date": "2019-06-03T03:38:55.146543600Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi, </p>\n\n<p>I am using CatBoost, but when I commit my code, there are so many files (like learn_err.tsv, meta.tsv,  test_error.tsv, time_left.tsv, catboost_training.json,  events.out.tfevents ...) in the output; currently I just manually choose the submission.csv and submit to test on public dataset.</p>\n\n<p>I am wondering if there are some methods that can prevent catboost generating irrelevant files?</p>\n\n<p>Thank you and good luck to all! </p>",
  "messages": [
    {
      "id": "541797",
      "postDate": "06/03/2019 03:38:55",
      "content": "<p>Hi, </p>\n\n<p>I am using CatBoost, but when I commit my code, there are so many files (like learn_err.tsv, meta.tsv,  test_error.tsv, time_left.tsv, catboost_training.json,  events.out.tfevents ...) in the output; currently I just manually choose the submission.csv and submit to test on public dataset.</p>\n\n<p>I am wondering if there are some methods that can prevent catboost generating irrelevant files?</p>\n\n<p>Thank you and good luck to all! </p>",
      "rawMarkdown": "Hi, \n\nI am using CatBoost, but when I commit my code, there are so many files (like learn_err.tsv, meta.tsv,  test_error.tsv, time_left.tsv, catboost_training.json,  events.out.tfevents ...) in the output; currently I just manually choose the submission.csv and submit to test on public dataset.\n\nI am wondering if there are some methods that can prevent catboost generating irrelevant files?\n\nThank you and good luck to all!",
      "votes": null
    },
    {
      "id": "541893",
      "postDate": "06/03/2019 07:02:06",
      "content": "<p>Just set <code>allow_writing_files=False</code> as documented <a href=\"https://catboost.ai/docs/concepts/python-reference_parameters-list.html\">here</a>.</p>",
      "rawMarkdown": "Just set `allow_writing_files=False` as documented [here](https://catboost.ai/docs/concepts/python-reference_parameters-list.html).",
      "votes": null
    },
    {
      "id": "542372",
      "postDate": "06/03/2019 19:14:21",
      "content": "<p>Thanks Berni! I am just afraid the other outputs of my kernel will mislead Kaggle to identify my submission.csv in the final submission (not daily 2 entries which I can manuelly select my file to submit)</p>",
      "rawMarkdown": "Thanks Berni! I am just afraid the other outputs of my kernel will mislead Kaggle to identify my submission.csv in the final submission (not daily 2 entries which I can manuelly select my file to submit)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 541893,
      "author_name": "bernir",
      "author_url": "",
      "post_date": "06/03/2019 07:02:06",
      "content": "<p>Just set <code>allow_writing_files=False</code> as documented <a href=\"https://catboost.ai/docs/concepts/python-reference_parameters-list.html\">here</a>.</p>",
      "votes": null,
      "replies": [
        {
          "id": 542372,
          "author_name": "zechengzhang1",
          "author_url": "",
          "post_date": "06/03/2019 19:14:21",
          "content": "<p>Thanks Berni! I am just afraid the other outputs of my kernel will mislead Kaggle to identify my submission.csv in the final submission (not daily 2 entries which I can manuelly select my file to submit)</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "541797": "Hi, \n\nI am using CatBoost, but when I commit my code, there are so many files (like learn_err.tsv, meta.tsv,  test_error.tsv, time_left.tsv, catboost_training.json,  events.out.tfevents ...) in the output; currently I just manually choose the submission.csv and submit to test on public dataset.\n\nI am wondering if there are some methods that can prevent catboost generating irrelevant files?\n\nThank you and good luck to all!",
    "541893": "Just set `allow_writing_files=False` as documented [here](https://catboost.ai/docs/concepts/python-reference_parameters-list.html).",
    "542372": "Thanks Berni! I am just afraid the other outputs of my kernel will mislead Kaggle to identify my submission.csv in the final submission (not daily 2 entries which I can manuelly select my file to submit)"
  },
  "source": "meta"
}