{ "id": 274930, "title": "Kaggle taking too long to commit for using TPU accelerator.", "url": "/competitions/g2net-gravitational-wave-detection/discussion/274930", "author_name": "", "post_date": "2021-09-28T07:01:30.853336800Z", "votes": 1, "comment_count": 6, "views": 0, "content": "

Hello All,
\nI used TPU as accelerator. I clicked on \"save and run all\" (commit) button kernel showing as committing…. from past 5 hours.
\nPlease provide solution. Thanks in advance.

\n

For Reference:
\nhttps://justpaste.it/2p1c1

", "messages": [ { "id": "1526512", "postDate": "09/28/2021 07:01:30", "content": "

Hello All,
\nI used TPU as accelerator. I clicked on \"save and run all\" (commit) button kernel showing as committing…. from past 5 hours.
\nPlease provide solution. Thanks in advance.

\n

For Reference:
\nhttps://justpaste.it/2p1c1

", "rawMarkdown": "Hello All,\nI used TPU as accelerator. I clicked on \"save and run all\" (commit) button kernel showing as committing.... from past 5 hours.\nPlease provide solution. Thanks in advance.\n\nFor Reference:\nhttps://justpaste.it/2p1c1", "votes": null }, { "id": "1526592", "postDate": "09/28/2021 08:11:30", "content": "

Kaggle is facing massive demand for TPUs; that's why we're all getting queued on resources.
\nYou can take a look at this discussion.

", "rawMarkdown": "Kaggle is facing massive demand for TPUs; that's why we're all getting queued on resources.\nYou can take a look at [this discussion](https://www.kaggle.com/c/g2net-gravitational-wave-detection/discussion/271623).", "votes": null }, { "id": "1526611", "postDate": "09/28/2021 08:28:16", "content": "

Thanks a lot.
\nIs network facility required through out the queue?
\nCan you tell approximate complete queue timing?

", "rawMarkdown": "Thanks a lot.\nIs network facility required through out the queue?\nCan you tell approximate complete queue timing?", "votes": null }, { "id": "1526644", "postDate": "09/28/2021 08:56:13", "content": "

No, you press \"Commit\" and wait till Kaggle launches your notebook.
\nYesterday I spent nearly 9 hours waiting for the start of notebook commit. My teammate spent 10+ hours.

", "rawMarkdown": "No, you press \"Commit\" and wait till Kaggle launches your notebook.\nYesterday I spent nearly 9 hours waiting for the start of notebook commit. My teammate spent 10+ hours.", "votes": null }, { "id": "1526666", "postDate": "09/28/2021 09:22:18", "content": "

This is just economics. There is a resource with limited supply, it is in demand. If everyone tries to use their TPU quota at once, there will be queues. No different from trying to drive to work in the morning peak. Given that Santa Claus isn't going to plug in a bunch of extra TPUs for free, the only options would be for Kaggle to further ration the resource (say cut everyone's TPU quota to 20h for the last week) or else to put up with the queues.

", "rawMarkdown": "This is just economics. There is a resource with limited supply, it is in demand. If everyone tries to use their TPU quota at once, there will be queues. No different from trying to drive to work in the morning peak. Given that Santa Claus isn't going to plug in a bunch of extra TPUs for free, the only options would be for Kaggle to further ration the resource (say cut everyone's TPU quota to 20h for the last week) or else to put up with the queues.", "votes": null }, { "id": "1526685", "postDate": "09/28/2021 09:35:02", "content": "

From my experience, most Kaggle competitions don't require extensive TPUs usage like this one. This competition is the first case in my memory when TPU demand is on fire.
\nThat is why I think rebalance is not needed. Of course, we would be happy to get more resources, yet TPUs are pretty expensive. GCP offers on-demand v3-8 TPUs at $8.00/hr, so we get $240 of TPU calculations each week.

", "rawMarkdown": "From my experience, most Kaggle competitions don't require extensive TPUs usage like this one. This competition is the first case in my memory when TPU demand is on fire.\nThat is why I think rebalance is not needed. Of course, we would be happy to get more resources, yet TPUs are pretty expensive. GCP offers on-demand v3-8 TPUs at $8.00/hr, so we get $240 of TPU calculations each week.", "votes": null }, { "id": "1559737", "postDate": "10/27/2021 07:10:33", "content": "

Hey All,

\n

Thank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.

\n

Thank you all,

\n

Regards,
\nChris

", "rawMarkdown": "Hey All,\n\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.\n\nThank you all,\n\nRegards,\nChris", "votes": null } ], "comments": [ { "id": 1526592, "author_name": "rolandw0w", "author_url": "", "post_date": "09/28/2021 08:11:30", "content": "

Kaggle is facing massive demand for TPUs; that's why we're all getting queued on resources.
\nYou can take a look at this discussion.

", "votes": null, "replies": [ { "id": 1526611, "author_name": "praveenthakkannavar", "author_url": "", "post_date": "09/28/2021 08:28:16", "content": "

Thanks a lot.
\nIs network facility required through out the queue?
\nCan you tell approximate complete queue timing?

", "votes": null, "replies": [] }, { "id": 1526644, "author_name": "rolandw0w", "author_url": "", "post_date": "09/28/2021 08:56:13", "content": "

No, you press \"Commit\" and wait till Kaggle launches your notebook.
\nYesterday I spent nearly 9 hours waiting for the start of notebook commit. My teammate spent 10+ hours.

", "votes": null, "replies": [] } ] }, { "id": 1526666, "author_name": "jbomitchell", "author_url": "", "post_date": "09/28/2021 09:22:18", "content": "

This is just economics. There is a resource with limited supply, it is in demand. If everyone tries to use their TPU quota at once, there will be queues. No different from trying to drive to work in the morning peak. Given that Santa Claus isn't going to plug in a bunch of extra TPUs for free, the only options would be for Kaggle to further ration the resource (say cut everyone's TPU quota to 20h for the last week) or else to put up with the queues.

", "votes": null, "replies": [ { "id": 1526685, "author_name": "rolandw0w", "author_url": "", "post_date": "09/28/2021 09:35:02", "content": "

From my experience, most Kaggle competitions don't require extensive TPUs usage like this one. This competition is the first case in my memory when TPU demand is on fire.
\nThat is why I think rebalance is not needed. Of course, we would be happy to get more resources, yet TPUs are pretty expensive. GCP offers on-demand v3-8 TPUs at $8.00/hr, so we get $240 of TPU calculations each week.

", "votes": null, "replies": [] } ] }, { "id": 1559737, "author_name": "zerafachris", "author_url": "", "post_date": "10/27/2021 07:10:33", "content": "

Hey All,

\n

Thank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.

\n

Thank you all,

\n

Regards,
\nChris

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1526512": "Hello All,\nI used TPU as accelerator. I clicked on \"save and run all\" (commit) button kernel showing as committing.... from past 5 hours.\nPlease provide solution. Thanks in advance.\n\nFor Reference:\nhttps://justpaste.it/2p1c1", "1526592": "Kaggle is facing massive demand for TPUs; that's why we're all getting queued on resources.\nYou can take a look at [this discussion](https://www.kaggle.com/c/g2net-gravitational-wave-detection/discussion/271623).", "1526611": "Thanks a lot.\nIs network facility required through out the queue?\nCan you tell approximate complete queue timing?", "1526644": "No, you press \"Commit\" and wait till Kaggle launches your notebook.\nYesterday I spent nearly 9 hours waiting for the start of notebook commit. My teammate spent 10+ hours.", "1526666": "This is just economics. There is a resource with limited supply, it is in demand. If everyone tries to use their TPU quota at once, there will be queues. No different from trying to drive to work in the morning peak. Given that Santa Claus isn't going to plug in a bunch of extra TPUs for free, the only options would be for Kaggle to further ration the resource (say cut everyone's TPU quota to 20h for the last week) or else to put up with the queues.", "1526685": "From my experience, most Kaggle competitions don't require extensive TPUs usage like this one. This competition is the first case in my memory when TPU demand is on fire.\nThat is why I think rebalance is not needed. Of course, we would be happy to get more resources, yet TPUs are pretty expensive. GCP offers on-demand v3-8 TPUs at $8.00/hr, so we get $240 of TPU calculations each week.", "1559737": "Hey All,\n\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.\n\nThank you all,\n\nRegards,\nChris" }, "source": "meta" }