{
  "id": 45840,
  "title": "Tensorflow issues",
  "url": "/competitions/tensorflow-speech-recognition-challenge/discussion/45840",
  "author_name": "Jagadeesh Dondeti",
  "post_date": "2017-12-16T10:56:33.445000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<ol>\n<li>Labels are getting shuffled</li>\n<li>When I am generating one hot encodings, the order is getting changed</li>\n<li>Probably, due to above issues my loss is not converging\nI am using queue runners while providing input to the network.</li>\n</ol>\n\n<p>If I am performing any sort of operation like subtraction of two tensors, then the outcome of the operation is totally shuffled.</p>\n\n<p>What could be the reason? Any kind of help would be appreciated.</p>",
  "messages": [
    {
      "id": 258531,
      "postDate": "2017-12-16T10:56:33.447Z",
      "content": "<ol>\n<li>Labels are getting shuffled</li>\n<li>When I am generating one hot encodings, the order is getting changed</li>\n<li>Probably, due to above issues my loss is not converging\nI am using queue runners while providing input to the network.</li>\n</ol>\n\n<p>If I am performing any sort of operation like subtraction of two tensors, then the outcome of the operation is totally shuffled.</p>\n\n<p>What could be the reason? Any kind of help would be appreciated.</p>",
      "rawMarkdown": " 1. Labels are getting shuffled\n 2. When I am generating one hot encodings, the order is getting changed\n 3. Probably, due to above issues my loss is not converging\nI am using queue runners while providing input to the network.\n\nIf I am performing any sort of operation like subtraction of two tensors, then the outcome of the operation is totally shuffled.\n    \n\nWhat could be the reason? Any kind of help would be appreciated.\n\n"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "258531": " 1. Labels are getting shuffled\n 2. When I am generating one hot encodings, the order is getting changed\n 3. Probably, due to above issues my loss is not converging\nI am using queue runners while providing input to the network.\n\nIf I am performing any sort of operation like subtraction of two tensors, then the outcome of the operation is totally shuffled.\n    \n\nWhat could be the reason? Any kind of help would be appreciated.\n\n"
  }
}