{
  "id": 71274,
  "title": "Request Fasttext .bin file instead of .vec",
  "url": "/competitions/quora-insincere-questions-classification/discussion/71274",
  "author_name": "",
  "post_date": "2018-11-12T05:58:55.395088100Z",
  "votes": 6,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Fasttext vectors and models are distributed in both a <strong>.bin</strong> file and a <strong>.vec</strong> file. While the <em>vec</em> file only contains the trained full-word vectors, the <em>bin</em> file contains the vectors as well as the Fasttext model with all of the sub-word ngram vectors. This is far more useful, as it allows us to leverage the full power of Fasttext and create decent embeddings for out-of-vocabulary words.</p>\n\n<p>Could we be provided the Fasttext <em>bin</em> file in addition to the <em>vec</em> file?</p>",
  "messages": [
    {
      "id": "419530",
      "postDate": "11/12/2018 05:58:55",
      "content": "<p>Fasttext vectors and models are distributed in both a <strong>.bin</strong> file and a <strong>.vec</strong> file. While the <em>vec</em> file only contains the trained full-word vectors, the <em>bin</em> file contains the vectors as well as the Fasttext model with all of the sub-word ngram vectors. This is far more useful, as it allows us to leverage the full power of Fasttext and create decent embeddings for out-of-vocabulary words.</p>\n\n<p>Could we be provided the Fasttext <em>bin</em> file in addition to the <em>vec</em> file?</p>",
      "rawMarkdown": "Fasttext vectors and models are distributed in both a **.bin** file and a **.vec** file. While the *vec* file only contains the trained full-word vectors, the *bin* file contains the vectors as well as the Fasttext model with all of the sub-word ngram vectors. This is far more useful, as it allows us to leverage the full power of Fasttext and create decent embeddings for out-of-vocabulary words.\n\nCould we be provided the Fasttext *bin* file in addition to the *vec* file?",
      "votes": null
    },
    {
      "id": "419780",
      "postDate": "11/12/2018 14:52:10",
      "content": "<p>Also, Fasttext trained on Common Crawl Data is much better !</p>",
      "rawMarkdown": "Also, Fasttext trained on Common Crawl Data is much better !",
      "votes": null
    },
    {
      "id": "420132",
      "postDate": "11/13/2018 05:32:40",
      "content": "<p>I request usage of deep emoji, sentimental lexicons, usage of twitter API, a cup of coffee and a new car</p>",
      "rawMarkdown": "I request usage of deep emoji, sentimental lexicons, usage of twitter API, a cup of coffee and a new car",
      "votes": null
    },
    {
      "id": "420156",
      "postDate": "11/13/2018 06:57:26",
      "content": "<p>nice share!</p>",
      "rawMarkdown": "nice share!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 419780,
      "author_name": "shaz13",
      "author_url": "",
      "post_date": "11/12/2018 14:52:10",
      "content": "<p>Also, Fasttext trained on Common Crawl Data is much better !</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 420132,
      "author_name": "christofhenkel",
      "author_url": "",
      "post_date": "11/13/2018 05:32:40",
      "content": "<p>I request usage of deep emoji, sentimental lexicons, usage of twitter API, a cup of coffee and a new car</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 420156,
      "author_name": "krystenlefebure",
      "author_url": "",
      "post_date": "11/13/2018 06:57:26",
      "content": "<p>nice share!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "419530": "Fasttext vectors and models are distributed in both a **.bin** file and a **.vec** file. While the *vec* file only contains the trained full-word vectors, the *bin* file contains the vectors as well as the Fasttext model with all of the sub-word ngram vectors. This is far more useful, as it allows us to leverage the full power of Fasttext and create decent embeddings for out-of-vocabulary words.\n\nCould we be provided the Fasttext *bin* file in addition to the *vec* file?",
    "419780": "Also, Fasttext trained on Common Crawl Data is much better !",
    "420132": "I request usage of deep emoji, sentimental lexicons, usage of twitter API, a cup of coffee and a new car",
    "420156": "nice share!"
  },
  "source": "meta"
}