{
  "id": 79876,
  "title": "Processing capacity - idea request",
  "url": "/competitions/elo-merchant-category-recommendation/discussion/79876",
  "author_name": "",
  "post_date": "2019-02-08T06:40:00.776799600Z",
  "votes": 1,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hi, I am trying to combine all the files together to build features. With my PC i5, and 8 GB ram the code is running for hours. \nIs there any suggestion to improve the processing speed. Thanks in advance!</p>",
  "messages": [
    {
      "id": "468023",
      "postDate": "02/08/2019 06:40:00",
      "content": "<p>Hi, I am trying to combine all the files together to build features. With my PC i5, and 8 GB ram the code is running for hours. \nIs there any suggestion to improve the processing speed. Thanks in advance!</p>",
      "rawMarkdown": "Hi, I am trying to combine all the files together to build features. With my PC i5, and 8 GB ram the code is running for hours. \nIs there any suggestion to improve the processing speed. Thanks in advance!",
      "votes": null
    },
    {
      "id": "468231",
      "postDate": "02/08/2019 14:43:29",
      "content": "<p>Why not work in a Kaggle kernel? It sounds like a kernel would be an upgrade from your local system.</p>",
      "rawMarkdown": "Why not work in a Kaggle kernel? It sounds like a kernel would be an upgrade from your local system.",
      "votes": null
    },
    {
      "id": "468502",
      "postDate": "02/09/2019 01:50:34",
      "content": "<p>Yes, it worked in Kaggle kernal. I have updated the script also. Thank you!</p>",
      "rawMarkdown": "Yes, it worked in Kaggle kernal. I have updated the script also. Thank you!",
      "votes": null
    },
    {
      "id": "469042",
      "postDate": "02/10/2019 11:16:57",
      "content": "<p>Hi Thomas,</p>\n\n<p>For small datasets/classic ML algos, I prefer my laptop (Thinkpad X1 carbon / 16GB RAM / i7-7600U processor). </p>\n\n<p>For any heavy dataset/Neural Network, I prefer Google Colaboratory and <a href=\"https://console.cloud.google.com\">GCP</a>. Google is offering free <a href=\"https://cloud.google.com/tpu/\">TPU</a> which is game changer for heavy computations!</p>\n\n<p>Here is a <a href=\"https://towardsdatascience.com/training-machine-learning-models-online-for-free-gpu-tpu-enabled-5def6a5c1ce3\">link</a>  in case you are new to it.</p>\n\n<p>HTH.</p>",
      "rawMarkdown": "Hi Thomas,\n\n\nFor small datasets/classic ML algos, I prefer my laptop (Thinkpad X1 carbon / 16GB RAM / i7-7600U processor). \n\nFor any heavy dataset/Neural Network, I prefer [Google Colaboratory](colab.research.google.com) and [GCP](https://console.cloud.google.com). Google is offering free [TPU](https://cloud.google.com/tpu/) which is game changer for heavy computations!\n\nHere is a [link](https://towardsdatascience.com/training-machine-learning-models-online-for-free-gpu-tpu-enabled-5def6a5c1ce3)  in case you are new to it.\n\nHTH.",
      "votes": null
    },
    {
      "id": "483069",
      "postDate": "03/04/2019 04:35:18",
      "content": "<p>Yes, Kaggle kernel is a wonderful alternative. Thanks!</p>",
      "rawMarkdown": "Yes, Kaggle kernel is a wonderful alternative. Thanks!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 468231,
      "author_name": "robikscube",
      "author_url": "",
      "post_date": "02/08/2019 14:43:29",
      "content": "<p>Why not work in a Kaggle kernel? It sounds like a kernel would be an upgrade from your local system.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 468502,
      "author_name": "mamkuzhy74",
      "author_url": "",
      "post_date": "02/09/2019 01:50:34",
      "content": "<p>Yes, it worked in Kaggle kernal. I have updated the script also. Thank you!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 469042,
      "author_name": "vjcalling",
      "author_url": "",
      "post_date": "02/10/2019 11:16:57",
      "content": "<p>Hi Thomas,</p>\n\n<p>For small datasets/classic ML algos, I prefer my laptop (Thinkpad X1 carbon / 16GB RAM / i7-7600U processor). </p>\n\n<p>For any heavy dataset/Neural Network, I prefer Google Colaboratory and <a href=\"https://console.cloud.google.com\">GCP</a>. Google is offering free <a href=\"https://cloud.google.com/tpu/\">TPU</a> which is game changer for heavy computations!</p>\n\n<p>Here is a <a href=\"https://towardsdatascience.com/training-machine-learning-models-online-for-free-gpu-tpu-enabled-5def6a5c1ce3\">link</a>  in case you are new to it.</p>\n\n<p>HTH.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 483069,
      "author_name": "mamkuzhy74",
      "author_url": "",
      "post_date": "03/04/2019 04:35:18",
      "content": "<p>Yes, Kaggle kernel is a wonderful alternative. Thanks!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "468023": "Hi, I am trying to combine all the files together to build features. With my PC i5, and 8 GB ram the code is running for hours. \nIs there any suggestion to improve the processing speed. Thanks in advance!",
    "468231": "Why not work in a Kaggle kernel? It sounds like a kernel would be an upgrade from your local system.",
    "468502": "Yes, it worked in Kaggle kernal. I have updated the script also. Thank you!",
    "469042": "Hi Thomas,\n\n\nFor small datasets/classic ML algos, I prefer my laptop (Thinkpad X1 carbon / 16GB RAM / i7-7600U processor). \n\nFor any heavy dataset/Neural Network, I prefer [Google Colaboratory](colab.research.google.com) and [GCP](https://console.cloud.google.com). Google is offering free [TPU](https://cloud.google.com/tpu/) which is game changer for heavy computations!\n\nHere is a [link](https://towardsdatascience.com/training-machine-learning-models-online-for-free-gpu-tpu-enabled-5def6a5c1ce3)  in case you are new to it.\n\nHTH.",
    "483069": "Yes, Kaggle kernel is a wonderful alternative. Thanks!"
  },
  "source": "meta"
}