Hello ,\nI would love to ask what kernel competition means ? And also do this kernels have time limit ? If so , what are the hardware configurations of a kernel?
", "messages": [ { "id": "422403", "postDate": "11/16/2018 07:10:51", "content": "Hello ,\nI would love to ask what kernel competition means ? And also do this kernels have time limit ? If so , what are the hardware configurations of a kernel?
", "rawMarkdown": "Hello ,\nI would love to ask what kernel competition means ? And also do this kernels have time limit ? If so , what are the hardware configurations of a kernel?", "votes": null }, { "id": "422412", "postDate": "11/16/2018 07:23:39", "content": "Kernel competition means you can submit only from Kaggle kernel. Run your kernel and then submit results.
\n\nTime limit is 6 hours, I think.
\n\nYou can configure options for your kernel, for example GPU on/off.
", "rawMarkdown": "Kernel competition means you can submit only from Kaggle kernel. Run your kernel and then submit results. \n\nTime limit is 6 hours, I think. \n\nYou can configure options for your kernel, for example GPU on/off.", "votes": null }, { "id": "422425", "postDate": "11/16/2018 07:48:29", "content": "I see , tnks for your rapid feedback , is it possible for me to make my models locally and just upload the result in kernel?
", "rawMarkdown": "I see , tnks for your rapid feedback , is it possible for me to make my models locally and just upload the result in kernel?", "votes": null }, { "id": "422479", "postDate": "11/16/2018 09:27:48", "content": "You can test your models locally but then your kernel needs to produce submission.csv based on Quora data and embeddings which are provided. See https://www.kaggle.com/c/quora-insincere-questions-classification#Kernels-FAQ
", "rawMarkdown": "You can test your models locally but then your kernel needs to produce submission.csv based on Quora data and embeddings which are provided. See https://www.kaggle.com/c/quora-insincere-questions-classification#Kernels-FAQ", "votes": null }, { "id": "422593", "postDate": "11/16/2018 13:06:20", "content": "Time limit for this competition is 6 hours if you don't activate the GPU, if you do it's 2 hours.\nThe GPU used is a Tesla K80 with 12Gb of memory.\nAnd you have 14 Gb of ram available here, but I believe it can vary from a competition to another.\nI don't know about the rest of the config, but It is not very important I guess.
", "rawMarkdown": "Time limit for this competition is 6 hours if you don't activate the GPU, if you do it's 2 hours.\nThe GPU used is a Tesla K80 with 12Gb of memory.\nAnd you have 14 Gb of ram available here, but I believe it can vary from a competition to another.\nI don't know about the rest of the config, but It is not very important I guess.", "votes": null } ], "comments": [ { "id": 422412, "author_name": "demonplus", "author_url": "", "post_date": "11/16/2018 07:23:39", "content": "Kernel competition means you can submit only from Kaggle kernel. Run your kernel and then submit results.
\n\nTime limit is 6 hours, I think.
\n\nYou can configure options for your kernel, for example GPU on/off.
", "votes": null, "replies": [ { "id": 422425, "author_name": "academycalbastard", "author_url": "", "post_date": "11/16/2018 07:48:29", "content": "I see , tnks for your rapid feedback , is it possible for me to make my models locally and just upload the result in kernel?
", "votes": null, "replies": [] }, { "id": 422479, "author_name": "demonplus", "author_url": "", "post_date": "11/16/2018 09:27:48", "content": "You can test your models locally but then your kernel needs to produce submission.csv based on Quora data and embeddings which are provided. See https://www.kaggle.com/c/quora-insincere-questions-classification#Kernels-FAQ
", "votes": null, "replies": [] } ] }, { "id": 422593, "author_name": "theoviel", "author_url": "", "post_date": "11/16/2018 13:06:20", "content": "Time limit for this competition is 6 hours if you don't activate the GPU, if you do it's 2 hours.\nThe GPU used is a Tesla K80 with 12Gb of memory.\nAnd you have 14 Gb of ram available here, but I believe it can vary from a competition to another.\nI don't know about the rest of the config, but It is not very important I guess.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "422403": "Hello ,\nI would love to ask what kernel competition means ? And also do this kernels have time limit ? If so , what are the hardware configurations of a kernel?", "422412": "Kernel competition means you can submit only from Kaggle kernel. Run your kernel and then submit results. \n\nTime limit is 6 hours, I think. \n\nYou can configure options for your kernel, for example GPU on/off.", "422425": "I see , tnks for your rapid feedback , is it possible for me to make my models locally and just upload the result in kernel?", "422479": "You can test your models locally but then your kernel needs to produce submission.csv based on Quora data and embeddings which are provided. See https://www.kaggle.com/c/quora-insincere-questions-classification#Kernels-FAQ", "422593": "Time limit for this competition is 6 hours if you don't activate the GPU, if you do it's 2 hours.\nThe GPU used is a Tesla K80 with 12Gb of memory.\nAnd you have 14 Gb of ram available here, but I believe it can vary from a competition to another.\nI don't know about the rest of the config, but It is not very important I guess." }, "source": "meta" }