{
  "id": 35228,
  "title": "Virtual Machine on Google Cloud for the datasets",
  "url": "/competitions/passenger-screening-algorithm-challenge/discussion/35228",
  "author_name": "",
  "post_date": "2017-06-24T16:55:42.497851900Z",
  "votes": 1,
  "comment_count": 4,
  "views": 1,
  "content": "<p>May I ask what is the VM name for all the datasets at Google cloud that one may launch from? </p>",
  "messages": [
    {
      "id": "195715",
      "postDate": "06/24/2017 16:55:42",
      "content": "<p>May I ask what is the VM name for all the datasets at Google cloud that one may launch from? </p>",
      "rawMarkdown": "May I ask what is the VM name for all the datasets at Google cloud that one may launch from?",
      "votes": null
    },
    {
      "id": "197562",
      "postDate": "06/29/2017 20:28:05",
      "content": "<p>Hi DrYe, </p>\n\n<p>The data is hosted in a <a href=\"https://cloud.google.com/storage/\">Google Cloud Storage</a> bucket not on a VM. You may find it useful to launch a VM using <a href=\"https://cloud.google.com/compute/\">Google Compute Engine</a>. </p>\n\n<p>Let me know if that helps clarify things. Thanks, - Steve</p>",
      "rawMarkdown": "Hi DrYe, \n\nThe data is hosted in a [Google Cloud Storage](https://cloud.google.com/storage/) bucket not on a VM. You may find it useful to launch a VM using [Google Compute Engine](https://cloud.google.com/compute/). \n\nLet me know if that helps clarify things. Thanks, - Steve",
      "votes": null
    },
    {
      "id": "198350",
      "postDate": "07/02/2017 01:47:57",
      "content": "<p>What GCE instance specs do you think are recommended for this dataset? </p>",
      "rawMarkdown": "What GCE instance specs do you think are recommended for this dataset?",
      "votes": null
    },
    {
      "id": "200037",
      "postDate": "07/06/2017 23:40:41",
      "content": "<p>Hi Jamie,</p>\n\n<p>That would also depend on the model architecture and algorithm you are using. I would suggest you try the Cloud ML Engine and look at its <a href=\"https://cloud.google.com/ml-engine/pricing#ml_training_units_by_scale_tier\">Scale Tiers</a>. You can start with BASIC and then increase the tier based on your needs.  </p>\n\n<p>Puneith</p>",
      "rawMarkdown": "Hi Jamie,\n\nThat would also depend on the model architecture and algorithm you are using. I would suggest you try the Cloud ML Engine and look at its [Scale Tiers][1]. You can start with BASIC and then increase the tier based on your needs.  \n\nPuneith\n\n\n  [1]: https://cloud.google.com/ml-engine/pricing#ml_training_units_by_scale_tier",
      "votes": null
    },
    {
      "id": "207522",
      "postDate": "07/26/2017 20:10:14",
      "content": "<p>This medium post - <a href=\"https://medium.com/google-cloud/jupyter-tensorflow-nvidia-gpu-docker-google-compute-engine-4a146f085f17\">Jupyter + Tensorflow + Nvidia GPU + Docker + Google Compute Engine</a> - gives some guidance about machine-sizing. </p>\n\n<p>(For the purposes of this image-intensive application, I'd recommending increasing the available RAM to the max for one vCPU - which I think is about 6 GB)</p>\n\n<p>Hope that helps. </p>",
      "rawMarkdown": "This medium post - [Jupyter + Tensorflow + Nvidia GPU + Docker + Google Compute Engine](https://medium.com/google-cloud/jupyter-tensorflow-nvidia-gpu-docker-google-compute-engine-4a146f085f17) - gives some guidance about machine-sizing. \n\n(For the purposes of this image-intensive application, I'd recommending increasing the available RAM to the max for one vCPU - which I think is about 6 GB)\n\nHope that helps.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 197562,
      "author_name": "stevegreenberg",
      "author_url": "",
      "post_date": "06/29/2017 20:28:05",
      "content": "<p>Hi DrYe, </p>\n\n<p>The data is hosted in a <a href=\"https://cloud.google.com/storage/\">Google Cloud Storage</a> bucket not on a VM. You may find it useful to launch a VM using <a href=\"https://cloud.google.com/compute/\">Google Compute Engine</a>. </p>\n\n<p>Let me know if that helps clarify things. Thanks, - Steve</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 198350,
      "author_name": "jameschartouni",
      "author_url": "",
      "post_date": "07/02/2017 01:47:57",
      "content": "<p>What GCE instance specs do you think are recommended for this dataset? </p>",
      "votes": null,
      "replies": [
        {
          "id": 207522,
          "author_name": "stevegreenberg",
          "author_url": "",
          "post_date": "07/26/2017 20:10:14",
          "content": "<p>This medium post - <a href=\"https://medium.com/google-cloud/jupyter-tensorflow-nvidia-gpu-docker-google-compute-engine-4a146f085f17\">Jupyter + Tensorflow + Nvidia GPU + Docker + Google Compute Engine</a> - gives some guidance about machine-sizing. </p>\n\n<p>(For the purposes of this image-intensive application, I'd recommending increasing the available RAM to the max for one vCPU - which I think is about 6 GB)</p>\n\n<p>Hope that helps. </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 200037,
      "author_name": "puneith",
      "author_url": "",
      "post_date": "07/06/2017 23:40:41",
      "content": "<p>Hi Jamie,</p>\n\n<p>That would also depend on the model architecture and algorithm you are using. I would suggest you try the Cloud ML Engine and look at its <a href=\"https://cloud.google.com/ml-engine/pricing#ml_training_units_by_scale_tier\">Scale Tiers</a>. You can start with BASIC and then increase the tier based on your needs.  </p>\n\n<p>Puneith</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "195715": "May I ask what is the VM name for all the datasets at Google cloud that one may launch from?",
    "197562": "Hi DrYe, \n\nThe data is hosted in a [Google Cloud Storage](https://cloud.google.com/storage/) bucket not on a VM. You may find it useful to launch a VM using [Google Compute Engine](https://cloud.google.com/compute/). \n\nLet me know if that helps clarify things. Thanks, - Steve",
    "198350": "What GCE instance specs do you think are recommended for this dataset?",
    "200037": "Hi Jamie,\n\nThat would also depend on the model architecture and algorithm you are using. I would suggest you try the Cloud ML Engine and look at its [Scale Tiers][1]. You can start with BASIC and then increase the tier based on your needs.  \n\nPuneith\n\n\n  [1]: https://cloud.google.com/ml-engine/pricing#ml_training_units_by_scale_tier",
    "207522": "This medium post - [Jupyter + Tensorflow + Nvidia GPU + Docker + Google Compute Engine](https://medium.com/google-cloud/jupyter-tensorflow-nvidia-gpu-docker-google-compute-engine-4a146f085f17) - gives some guidance about machine-sizing. \n\n(For the purposes of this image-intensive application, I'd recommending increasing the available RAM to the max for one vCPU - which I think is about 6 GB)\n\nHope that helps."
  },
  "source": "meta"
}