{
  "fetched_at": "2026-10-19T07:12:24Z",
  "source_url": "https://www.kaggle.com/competitions/11785-hw2/data",
  "pages": [
    {
      "name": "rules",
      "content": "CMU University Policy on Academic Integrity applies. Submit your own work and your own work only unless otherwise specified. Try to have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "data-description",
      "content": "#Data Files\n\nData files are available for download from this page.\n\n##File descriptions\n\n* **train-features.npy** - training features\n* **train-labels.npy** - training labels\n* **dev-features.npy** -  development/validation features\n* **dev-labels.npy** -  development/validation labels\n* **test-features.npy** - test features\n* **sample_submission.csv** - sample submission in correct format\n\n##Feature Files\n\n`[train|dev|test]-features.npy` contain a numpy object array of shape `[samples, 2]`. The first column contains feature arrays, the second column contains bound arrays. Each row is the features and bounds of a single utterance, that may contain multiple phonemes.\n\n```\ndata = np.load('train-features.npy')\n```\n\n- **`data[i, 0]` (features)** - float32 numpy array of shape `[time, frequency]`. Frequency dimension is always 40 but time dimension is of variable length.\n- **`data[i, 1]` (bounds)** - int32 numpy array of shape `[phoneme count]`. Each value indicates index (0-based) at which one phoneme begins and previous phoneme ends. Index refers to time dimension in feature array. The length of this array is the number of phonemes in that utterance.\n\nThe features of phoneme `i` in utterance `j` are `data[j,0][data[j,1][i]:data[j,1][i+1]]`, unless `i+1==data[j,1].shape[0]`, in which case the features are `data[j,0][data[j,1][i]:]`.\n\n##Label Files\n\n`[train|dev]-labels.npy` contain a numpy object array of shape `[samples]`. Each element in the array is an int32 array of shape `[phoneme count]` and provides the label for each corresponding phoneme. There are 46 output labels [0-45].",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "This is not the competition you are looking for. Please go to the updated competition (link below).\n\nhttps://www.kaggle.com/c/11785-hw2pt2",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}