{
  "id": 240829,
  "title": "GPU Crisis",
  "url": "/competitions/bms-molecular-translation/discussion/240829",
  "author_name": "",
  "post_date": "2021-05-21T14:58:50.792001200Z",
  "votes": 22,
  "comment_count": 13,
  "views": 0,
  "content": "<p>I am having 1000 usd GCP credits , but I am unable to increase my GPU quota . The requests are getting rejected and I am using my personal account and not organizational account . Has anybody else faced this issue ? I am based out of India .</p>\n<p>Currently I am using GPU credits from Jarvislabs.ai (<a href=\"https://cloud.jarvislabs.ai/\" target=\"_blank\">https://cloud.jarvislabs.ai/</a>) . I find Jarvislabs.ai  more affordable than other Cloud providers and also there is no issue of quota increase since it's prepaid . I am also able to get A100s without any hassles.</p>",
  "messages": [
    {
      "id": "1317626",
      "postDate": "05/21/2021 14:58:50",
      "content": "<p>I am having 1000 usd GCP credits , but I am unable to increase my GPU quota . The requests are getting rejected and I am using my personal account and not organizational account . Has anybody else faced this issue ? I am based out of India .</p>\n<p>Currently I am using GPU credits from Jarvislabs.ai (<a href=\"https://cloud.jarvislabs.ai/\" target=\"_blank\">https://cloud.jarvislabs.ai/</a>) . I find Jarvislabs.ai  more affordable than other Cloud providers and also there is no issue of quota increase since it's prepaid . I am also able to get A100s without any hassles.</p>",
      "rawMarkdown": "I am having 1000 usd GCP credits , but I am unable to increase my GPU quota . The requests are getting rejected and I am using my personal account and not organizational account . Has anybody else faced this issue ? I am based out of India .\n\nCurrently I am using GPU credits from Jarvislabs.ai (https://cloud.jarvislabs.ai/) . I find Jarvislabs.ai  more affordable than other Cloud providers and also there is no issue of quota increase since it's prepaid . I am also able to get A100s without any hassles.",
      "votes": null
    },
    {
      "id": "1317801",
      "postDate": "05/21/2021 17:56:01",
      "content": "<p>Been there <a href=\"https://www.kaggle.com/usharengaraju\" target=\"_blank\">@usharengaraju</a> :( . I had access to 1000$ too. But didn't get GPU access. It's sad that accessing <strong>GCP</strong> GPU is very difficult whereas you can access <strong>AWS</strong> GPU very easily.</p>",
      "rawMarkdown": "Been there @usharengaraju :( . I had access to 1000$ too. But didn't get GPU access. It's sad that accessing **GCP** GPU is very difficult whereas you can access **AWS** GPU very easily.",
      "votes": null
    },
    {
      "id": "1317806",
      "postDate": "05/21/2021 17:58:20",
      "content": "<p><a href=\"https://www.kaggle.com/awsaf49\" target=\"_blank\">@awsaf49</a> Totally agree </p>",
      "rawMarkdown": "awsaf49 Totally agree",
      "votes": null
    },
    {
      "id": "1317924",
      "postDate": "05/21/2021 20:01:23",
      "content": "<p>I've had difficulties, through work, when attempting to access GPUs in certain regions (Central specifically). Other regions seem much better. Can you try other regions?</p>\n<p>Regardless, it's incredibly frustrating. I hope they figure it out and you can get the access!</p>",
      "rawMarkdown": "I've had difficulties, through work, when attempting to access GPUs in certain regions (Central specifically). Other regions seem much better. Can you try other regions?\n\nRegardless, it's incredibly frustrating. I hope they figure it out and you can get the access!",
      "votes": null
    },
    {
      "id": "1318292",
      "postDate": "05/22/2021 07:34:46",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/usharengaraju\" target=\"_blank\">@usharengaraju</a> , <a href=\"https://www.kaggle.com/dschettler8845\" target=\"_blank\">@dschettler8845</a> <a href=\"https://www.kaggle.com/awsaf49\" target=\"_blank\">@awsaf49</a> :)</p>\n<p>from my experience, I don't recommend using a personal account to increase any quota in GCP.</p>\n<p>The best way I have found is to ask your organization to create an email account called </p>\n<p>data.scientists@……</p>\n<p>and with a detailed explanation of why you want to increase the quota, everything will go smoothly. </p>",
      "rawMarkdown": "Hi @usharengaraju , @dschettler8845 @awsaf49 :)\n\nfrom my experience, I don't recommend using a personal account to increase any quota in GCP.\n\nThe best way I have found is to ask your organization to create an email account called \n\ndata.scientists@......\n\nand with a detailed explanation of why you want to increase the quota, everything will go smoothly.",
      "votes": null
    },
    {
      "id": "1318341",
      "postDate": "05/22/2021 08:27:12",
      "content": "<p>I couldn't increase my quota either the last time I tried. </p>\n<p>I  can only use single GPU A100/V100</p>",
      "rawMarkdown": "I couldn't increase my quota either the last time I tried. \n\nI  can only use single GPU A100/V100",
      "votes": null
    },
    {
      "id": "1318367",
      "postDate": "05/22/2021 09:08:11",
      "content": "<p>haha, I couldn't even access a single GPU let alone multiple. And here I was thinking that Google had a good <strong>marketing team</strong> and tools to analyze people XD.</p>",
      "rawMarkdown": "haha, I couldn't even access a single GPU let alone multiple. And here I was thinking that Google had a good **marketing team** and tools to analyze people XD.",
      "votes": null
    },
    {
      "id": "1318404",
      "postDate": "05/22/2021 10:03:19",
      "content": "<p>You can send a request for single A100 GPU with personnal account,  multi GPU is more complicated. </p>",
      "rawMarkdown": "You can send a request for single A100 GPU with personnal account,  multi GPU is more complicated.",
      "votes": null
    },
    {
      "id": "1318407",
      "postDate": "05/22/2021 10:06:01",
      "content": "<p>they reject saying I don't have any prior history to that project …</p>",
      "rawMarkdown": "they reject saying I don't have any prior history to that project ...",
      "votes": null
    },
    {
      "id": "1321575",
      "postDate": "05/24/2021 19:09:42",
      "content": "<p>If your training time is less than equal to 24 hours than try colab pro. It just 10 USD per month and now available in India.</p>",
      "rawMarkdown": "If your training time is less than equal to 24 hours than try colab pro. It just 10 USD per month and now available in India.",
      "votes": null
    },
    {
      "id": "1321703",
      "postDate": "05/24/2021 22:00:18",
      "content": "<p>I like colab pro.</p>",
      "rawMarkdown": "I like colab pro.",
      "votes": null
    },
    {
      "id": "1322075",
      "postDate": "05/25/2021 07:16:15",
      "content": "<p>Is there any effective way to transfer the data into google drive? If we try using kaggle's cli, it creates a zip file and extracting it takes a long time. </p>",
      "rawMarkdown": "Is there any effective way to transfer the data into google drive? If we try using kaggle's cli, it creates a zip file and extracting it takes a long time.",
      "votes": null
    },
    {
      "id": "1322426",
      "postDate": "05/25/2021 12:55:58",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/tamilselvanmoorthy\" target=\"_blank\">@tamilselvanmoorthy</a> :)</p>\n<p>See this (<strong>The pros and cons of the 3 options to deal with Kaggle Datasets in Colab</strong> )<br>\n<a href=\"https://www.kaggle.com/c/birdclef-2021/discussion/241414\" target=\"_blank\">https://www.kaggle.com/c/birdclef-2021/discussion/241414</a></p>",
      "rawMarkdown": "Hi @tamilselvanmoorthy :)\n\nSee this (**The pros and cons of the 3 options to deal with Kaggle Datasets in Colab** )\nhttps://www.kaggle.com/c/birdclef-2021/discussion/241414",
      "votes": null
    },
    {
      "id": "1322874",
      "postDate": "05/25/2021 18:51:50",
      "content": "<p><a href=\"https://www.kaggle.com/tamilselvanmoorthy\" target=\"_blank\">@tamilselvanmoorthy</a> did u try this ??<br>\n<a href=\"https://www.kaggle.com/general/74235\" target=\"_blank\">https://www.kaggle.com/general/74235</a></p>",
      "rawMarkdown": "tamilselvanmoorthy did u try this ??\nhttps://www.kaggle.com/general/74235",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1317801,
      "author_name": "awsaf49",
      "author_url": "",
      "post_date": "05/21/2021 17:56:01",
      "content": "<p>Been there <a href=\"https://www.kaggle.com/usharengaraju\" target=\"_blank\">@usharengaraju</a> :( . I had access to 1000$ too. But didn't get GPU access. It's sad that accessing <strong>GCP</strong> GPU is very difficult whereas you can access <strong>AWS</strong> GPU very easily.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1317806,
          "author_name": "usharengaraju",
          "author_url": "",
          "post_date": "05/21/2021 17:58:20",
          "content": "<p><a href=\"https://www.kaggle.com/awsaf49\" target=\"_blank\">@awsaf49</a> Totally agree </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1317924,
      "author_name": "dschettler8845",
      "author_url": "",
      "post_date": "05/21/2021 20:01:23",
      "content": "<p>I've had difficulties, through work, when attempting to access GPUs in certain regions (Central specifically). Other regions seem much better. Can you try other regions?</p>\n<p>Regardless, it's incredibly frustrating. I hope they figure it out and you can get the access!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1318292,
      "author_name": "faisalalsrheed",
      "author_url": "",
      "post_date": "05/22/2021 07:34:46",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/usharengaraju\" target=\"_blank\">@usharengaraju</a> , <a href=\"https://www.kaggle.com/dschettler8845\" target=\"_blank\">@dschettler8845</a> <a href=\"https://www.kaggle.com/awsaf49\" target=\"_blank\">@awsaf49</a> :)</p>\n<p>from my experience, I don't recommend using a personal account to increase any quota in GCP.</p>\n<p>The best way I have found is to ask your organization to create an email account called </p>\n<p>data.scientists@……</p>\n<p>and with a detailed explanation of why you want to increase the quota, everything will go smoothly. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1318341,
      "author_name": "serigne",
      "author_url": "",
      "post_date": "05/22/2021 08:27:12",
      "content": "<p>I couldn't increase my quota either the last time I tried. </p>\n<p>I  can only use single GPU A100/V100</p>",
      "votes": null,
      "replies": [
        {
          "id": 1318367,
          "author_name": "awsaf49",
          "author_url": "",
          "post_date": "05/22/2021 09:08:11",
          "content": "<p>haha, I couldn't even access a single GPU let alone multiple. And here I was thinking that Google had a good <strong>marketing team</strong> and tools to analyze people XD.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1318404,
          "author_name": "serigne",
          "author_url": "",
          "post_date": "05/22/2021 10:03:19",
          "content": "<p>You can send a request for single A100 GPU with personnal account,  multi GPU is more complicated. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1318407,
          "author_name": "awsaf49",
          "author_url": "",
          "post_date": "05/22/2021 10:06:01",
          "content": "<p>they reject saying I don't have any prior history to that project …</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1321575,
      "author_name": "vikrant06",
      "author_url": "",
      "post_date": "05/24/2021 19:09:42",
      "content": "<p>If your training time is less than equal to 24 hours than try colab pro. It just 10 USD per month and now available in India.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1321703,
          "author_name": "faisalalsrheed",
          "author_url": "",
          "post_date": "05/24/2021 22:00:18",
          "content": "<p>I like colab pro.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1322075,
          "author_name": "tamilselvanmoorthy",
          "author_url": "",
          "post_date": "05/25/2021 07:16:15",
          "content": "<p>Is there any effective way to transfer the data into google drive? If we try using kaggle's cli, it creates a zip file and extracting it takes a long time. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1322426,
          "author_name": "faisalalsrheed",
          "author_url": "",
          "post_date": "05/25/2021 12:55:58",
          "content": "<p>Hi <a href=\"https://www.kaggle.com/tamilselvanmoorthy\" target=\"_blank\">@tamilselvanmoorthy</a> :)</p>\n<p>See this (<strong>The pros and cons of the 3 options to deal with Kaggle Datasets in Colab</strong> )<br>\n<a href=\"https://www.kaggle.com/c/birdclef-2021/discussion/241414\" target=\"_blank\">https://www.kaggle.com/c/birdclef-2021/discussion/241414</a></p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1322874,
          "author_name": "vikrant06",
          "author_url": "",
          "post_date": "05/25/2021 18:51:50",
          "content": "<p><a href=\"https://www.kaggle.com/tamilselvanmoorthy\" target=\"_blank\">@tamilselvanmoorthy</a> did u try this ??<br>\n<a href=\"https://www.kaggle.com/general/74235\" target=\"_blank\">https://www.kaggle.com/general/74235</a></p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1317626": "I am having 1000 usd GCP credits , but I am unable to increase my GPU quota . The requests are getting rejected and I am using my personal account and not organizational account . Has anybody else faced this issue ? I am based out of India .\n\nCurrently I am using GPU credits from Jarvislabs.ai (https://cloud.jarvislabs.ai/) . I find Jarvislabs.ai  more affordable than other Cloud providers and also there is no issue of quota increase since it's prepaid . I am also able to get A100s without any hassles.",
    "1317801": "Been there @usharengaraju :( . I had access to 1000$ too. But didn't get GPU access. It's sad that accessing **GCP** GPU is very difficult whereas you can access **AWS** GPU very easily.",
    "1317806": "awsaf49 Totally agree",
    "1317924": "I've had difficulties, through work, when attempting to access GPUs in certain regions (Central specifically). Other regions seem much better. Can you try other regions?\n\nRegardless, it's incredibly frustrating. I hope they figure it out and you can get the access!",
    "1318292": "Hi @usharengaraju , @dschettler8845 @awsaf49 :)\n\nfrom my experience, I don't recommend using a personal account to increase any quota in GCP.\n\nThe best way I have found is to ask your organization to create an email account called \n\ndata.scientists@......\n\nand with a detailed explanation of why you want to increase the quota, everything will go smoothly.",
    "1318341": "I couldn't increase my quota either the last time I tried. \n\nI  can only use single GPU A100/V100",
    "1318367": "haha, I couldn't even access a single GPU let alone multiple. And here I was thinking that Google had a good **marketing team** and tools to analyze people XD.",
    "1318404": "You can send a request for single A100 GPU with personnal account,  multi GPU is more complicated.",
    "1318407": "they reject saying I don't have any prior history to that project ...",
    "1321575": "If your training time is less than equal to 24 hours than try colab pro. It just 10 USD per month and now available in India.",
    "1321703": "I like colab pro.",
    "1322075": "Is there any effective way to transfer the data into google drive? If we try using kaggle's cli, it creates a zip file and extracting it takes a long time.",
    "1322426": "Hi @tamilselvanmoorthy :)\n\nSee this (**The pros and cons of the 3 options to deal with Kaggle Datasets in Colab** )\nhttps://www.kaggle.com/c/birdclef-2021/discussion/241414",
    "1322874": "tamilselvanmoorthy did u try this ??\nhttps://www.kaggle.com/general/74235"
  },
  "source": "meta"
}