{
  "id": 18688,
  "title": "Applying existing Conv nets in Caffe with two GPUs.",
  "url": "/competitions/second-annual-data-science-bowl/discussion/18688",
  "author_name": "blazingshrooms",
  "post_date": "2016-02-01T19:03:55.583000",
  "votes": 0,
  "comment_count": 0,
  "views": 464,
  "content": "<p>I have a conv net that I trained using Caffe. When training, Caffe supported using multiple GPUs to compute efficiently. I was wondering if there is a similar configuration that I can use when I am applying my conv net to the testing data.</p>\n\n<p>Thanks!</p>",
  "messages": [
    {
      "id": 106522,
      "postDate": "2016-02-01T19:03:55.583Z",
      "content": "<p>I have a conv net that I trained using Caffe. When training, Caffe supported using multiple GPUs to compute efficiently. I was wondering if there is a similar configuration that I can use when I am applying my conv net to the testing data.</p>\n\n<p>Thanks!</p>",
      "rawMarkdown": "I have a conv net that I trained using Caffe. When training, Caffe supported using multiple GPUs to compute efficiently. I was wondering if there is a similar configuration that I can use when I am applying my conv net to the testing data.\r\n\r\nThanks!"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "106522": "I have a conv net that I trained using Caffe. When training, Caffe supported using multiple GPUs to compute efficiently. I was wondering if there is a similar configuration that I can use when I am applying my conv net to the testing data.\r\n\r\nThanks!"
  }
}