New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!
\n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!
\n\nNew to Kaggle? Take a look at a few videos our very own Dr. Rachael Tatman has put together to learn a bit more about site etiquette, Kaggle lingo, and how to enter a competition using Kaggle Notebooks.
", "messages": [ { "id": "660251", "postDate": "10/28/2019 23:46:36", "content": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!
\n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!
\n\nNew to Kaggle? Take a look at a few videos our very own Dr. Rachael Tatman has put together to learn a bit more about site etiquette, Kaggle lingo, and how to enter a competition using Kaggle Notebooks.
", "rawMarkdown": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with! \n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!\n\nNew to Kaggle? Take a look at a few videos our very own Dr. Rachael Tatman has put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s), and [how to enter a competition using Kaggle Notebooks](https://www.youtube.com/watch?&v=GJBOMWpLpTQ).", "votes": null }, { "id": "664079", "postDate": "11/03/2019 05:55:48", "content": "Do we need to use tensorflow 2.0 or we are free to use any other framework?\nAlso how we can upload the trained model?
", "rawMarkdown": "Do we need to use tensorflow 2.0 or we are free to use any other framework?\nAlso how we can upload the trained model?", "votes": null }, { "id": "669677", "postDate": "11/10/2019 10:14:45", "content": "Answer is in another thread: https://www.kaggle.com/c/tensorflow2-question-answering/discussion/114727#latest-665122
", "rawMarkdown": "Answer is in another thread: https://www.kaggle.com/c/tensorflow2-question-answering/discussion/114727#latest-665122", "votes": null }, { "id": "676724", "postDate": "11/19/2019 13:12:22", "content": "I'm learning machine learning.\nI've been studying for nearly a year, but think of it as a beginner.\nkaggle performed 3 times. (MNIST, Titanic, House Prices)\nI'm still not enough to study, but I want to challenge \"Featured Code Competition\".\nHow should I prepare?\nPlease teach me.
", "rawMarkdown": "I'm learning machine learning.\nI've been studying for nearly a year, but think of it as a beginner.\nkaggle performed 3 times. (MNIST, Titanic, House Prices)\nI'm still not enough to study, but I want to challenge \"Featured Code Competition\".\nHow should I prepare?\nPlease teach me.", "votes": null }, { "id": "681505", "postDate": "11/26/2019 07:07:38", "content": "Why cannot we use deeppavlov?
\n\npython -m deeppavlov install squad_bert\nfrom deeppavlov import build_model, configs\nmodel = build_model(configs.squad.squad, download=True)\nmodel(['DeepPavlov is library for NLP and dialog systems.'], ['What is DeepPavlov?'])
\n\nAlso, is this a kernels only competition?
", "rawMarkdown": "Why cannot we use deeppavlov?\n\npython -m deeppavlov install squad_bert\nfrom deeppavlov import build_model, configs\nmodel = build_model(configs.squad.squad, download=True)\nmodel(['DeepPavlov is library for NLP and dialog systems.'], ['What is DeepPavlov?'])\n\nAlso, is this a kernels only competition?", "votes": null }, { "id": "687255", "postDate": "12/04/2019 05:59:38", "content": "How does Kaggle run your notebook (in regards to the input test set) to determine the final winners?\nAnd how must you name your input files/datasets?\nDoes the \"public\" input file have the same name as the \"private\" input file?
\n\nLet me give some background and explain my question in detail: (it's a very simple question but it's hard to explain - thus long-winded)\nI'm trying to understand the mechanics of how your notebook (or kernel as I guess they are called) gets run by kaggle to determine \"private leaderboard\" and the ultimate winners of competition.\nIn particular since your notebook will have an input file with a name like:\n'/kaggle/input/tensorflow2-question-answering/simplified-nq-test.jsonl'\nthen how do they run your notebook against the \"private\" test data?\nThe competition rules - description says that they'll run your notebook against a \"private\" data test set.\nSo does this private data set have the same name as the test set they give you?\nDo they simply swap out the 'simplified-nq-test.jsonl' file that they give to each competitor for their own\nsecret data file with the same name 'simplified-nq-test.jsonl'?\nOr is the \"secret\" data file have a different name?\nIf it's different then that means your notebook code must account for that.
\n\nThis is really confusing for 1st time competitors as I could find no explanation of this on the competition web page.\nBut when I created a notebook by branching off of a public notebook the above input file name was used within the public notebook.\nThus I can only conclude that is the proper name and the proper subdirectory (data-set) name.\nBut if I'm wrong on this point then my notebook will surely fail when the kaggle-competition committee runs it to determine private leaderboard ranking.
\n\nThe whole point of this question is the competitors MUST know how to name the input file within their notebook.
", "rawMarkdown": "How does Kaggle run your notebook (in regards to the input test set) to determine the final winners?\nAnd how must you name your input files/datasets?\nDoes the \"public\" input file have the same name as the \"private\" input file?\n\nLet me give some background and explain my question in detail: (it's a very simple question but it's hard to explain - thus long-winded)\nI'm trying to understand the mechanics of how your notebook (or kernel as I guess they are called) gets run by kaggle to determine \"private leaderboard\" and the ultimate winners of competition.\nIn particular since your notebook will have an input file with a name like:\n'/kaggle/input/tensorflow2-question-answering/simplified-nq-test.jsonl'\nthen how do they run your notebook against the \"private\" test data?\nThe competition rules - description says that they'll run your notebook against a \"private\" data test set.\nSo does this private data set have the same name as the test set they give you?\nDo they simply swap out the 'simplified-nq-test.jsonl' file that they give to each competitor for their own\nsecret data file with the same name 'simplified-nq-test.jsonl'?\nOr is the \"secret\" data file have a different name?\nIf it's different then that means your notebook code must account for that.\n\nThis is really confusing for 1st time competitors as I could find no explanation of this on the competition web page.\nBut when I created a notebook by branching off of a public notebook the above input file name was used within the public notebook.\nThus I can only conclude that is the proper name and the proper subdirectory (data-set) name.\nBut if I'm wrong on this point then my notebook will surely fail when the kaggle-competition committee runs it to determine private leaderboard ranking.\n\nThe whole point of this question is the competitors MUST know how to name the input file within their notebook.", "votes": null }, { "id": "687307", "postDate": "12/04/2019 08:00:14", "content": "I agree that it's not well explained currently, I can't find a source to link to, except for http://blog.kaggle.com/2019/05/17/synchronous-kernels-only-competitions-real-ml-in-real-time/, but this is my understanding:\n- on each submission, the notebook is run twice, first against test set which we can download on the data page, second on unseen private test set\n- file names are the same, only the content is swapped\n- since the notebook is run against both sets, you can be sure that your notebook didn't throw an error when running against private set (if your submission succeeded). But still be sure to make your code perform well on the file you don't see.\n- to avoid probing the test set, you don't see output from your notebook when it's run on private test, and if it's an error, you don't get an error message
", "rawMarkdown": "I agree that it's not well explained currently, I can't find a source to link to, except for http://blog.kaggle.com/2019/05/17/synchronous-kernels-only-competitions-real-ml-in-real-time/, but this is my understanding:\n- on each submission, the notebook is run twice, first against test set which we can download on the data page, second on unseen private test set\n- file names are the same, only the content is swapped\n- since the notebook is run against both sets, you can be sure that your notebook didn't throw an error when running against private set (if your submission succeeded). But still be sure to make your code perform well on the file you don't see.\n- to avoid probing the test set, you don't see output from your notebook when it's run on private test, and if it's an error, you don't get an error message", "votes": null }, { "id": "694569", "postDate": "12/13/2019 19:32:14", "content": "Are most notebooks using GCP for training? Mine took 4 hours for 8 samples!
", "rawMarkdown": "Are most notebooks using GCP for training? Mine took 4 hours for 8 samples!", "votes": null }, { "id": "695395", "postDate": "12/15/2019 04:03:46", "content": "Did you enable the GPU? Just checking...
", "rawMarkdown": "Did you enable the GPU? Just checking...", "votes": null } ], "comments": [ { "id": 664079, "author_name": "atulkum", "author_url": "", "post_date": "11/03/2019 05:55:48", "content": "Do we need to use tensorflow 2.0 or we are free to use any other framework?\nAlso how we can upload the trained model?
", "votes": null, "replies": [ { "id": 669677, "author_name": "arvindpdmn", "author_url": "", "post_date": "11/10/2019 10:14:45", "content": "Answer is in another thread: https://www.kaggle.com/c/tensorflow2-question-answering/discussion/114727#latest-665122
", "votes": null, "replies": [] } ] }, { "id": 676724, "author_name": "matrix99999", "author_url": "", "post_date": "11/19/2019 13:12:22", "content": "I'm learning machine learning.\nI've been studying for nearly a year, but think of it as a beginner.\nkaggle performed 3 times. (MNIST, Titanic, House Prices)\nI'm still not enough to study, but I want to challenge \"Featured Code Competition\".\nHow should I prepare?\nPlease teach me.
", "votes": null, "replies": [] }, { "id": 681505, "author_name": "isikkuntay", "author_url": "", "post_date": "11/26/2019 07:07:38", "content": "Why cannot we use deeppavlov?
\n\npython -m deeppavlov install squad_bert\nfrom deeppavlov import build_model, configs\nmodel = build_model(configs.squad.squad, download=True)\nmodel(['DeepPavlov is library for NLP and dialog systems.'], ['What is DeepPavlov?'])
\n\nAlso, is this a kernels only competition?
", "votes": null, "replies": [] }, { "id": 687255, "author_name": "mchnln", "author_url": "", "post_date": "12/04/2019 05:59:38", "content": "How does Kaggle run your notebook (in regards to the input test set) to determine the final winners?\nAnd how must you name your input files/datasets?\nDoes the \"public\" input file have the same name as the \"private\" input file?
\n\nLet me give some background and explain my question in detail: (it's a very simple question but it's hard to explain - thus long-winded)\nI'm trying to understand the mechanics of how your notebook (or kernel as I guess they are called) gets run by kaggle to determine \"private leaderboard\" and the ultimate winners of competition.\nIn particular since your notebook will have an input file with a name like:\n'/kaggle/input/tensorflow2-question-answering/simplified-nq-test.jsonl'\nthen how do they run your notebook against the \"private\" test data?\nThe competition rules - description says that they'll run your notebook against a \"private\" data test set.\nSo does this private data set have the same name as the test set they give you?\nDo they simply swap out the 'simplified-nq-test.jsonl' file that they give to each competitor for their own\nsecret data file with the same name 'simplified-nq-test.jsonl'?\nOr is the \"secret\" data file have a different name?\nIf it's different then that means your notebook code must account for that.
\n\nThis is really confusing for 1st time competitors as I could find no explanation of this on the competition web page.\nBut when I created a notebook by branching off of a public notebook the above input file name was used within the public notebook.\nThus I can only conclude that is the proper name and the proper subdirectory (data-set) name.\nBut if I'm wrong on this point then my notebook will surely fail when the kaggle-competition committee runs it to determine private leaderboard ranking.
\n\nThe whole point of this question is the competitors MUST know how to name the input file within their notebook.
", "votes": null, "replies": [ { "id": 687307, "author_name": "lopuhin", "author_url": "", "post_date": "12/04/2019 08:00:14", "content": "I agree that it's not well explained currently, I can't find a source to link to, except for http://blog.kaggle.com/2019/05/17/synchronous-kernels-only-competitions-real-ml-in-real-time/, but this is my understanding:\n- on each submission, the notebook is run twice, first against test set which we can download on the data page, second on unseen private test set\n- file names are the same, only the content is swapped\n- since the notebook is run against both sets, you can be sure that your notebook didn't throw an error when running against private set (if your submission succeeded). But still be sure to make your code perform well on the file you don't see.\n- to avoid probing the test set, you don't see output from your notebook when it's run on private test, and if it's an error, you don't get an error message
", "votes": null, "replies": [] } ] }, { "id": 694569, "author_name": "isikkuntay", "author_url": "", "post_date": "12/13/2019 19:32:14", "content": "Are most notebooks using GCP for training? Mine took 4 hours for 8 samples!
", "votes": null, "replies": [ { "id": 695395, "author_name": "arvindpdmn", "author_url": "", "post_date": "12/15/2019 04:03:46", "content": "Did you enable the GPU? Just checking...
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "660251": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with! \n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!\n\nNew to Kaggle? Take a look at a few videos our very own Dr. Rachael Tatman has put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s), and [how to enter a competition using Kaggle Notebooks](https://www.youtube.com/watch?&v=GJBOMWpLpTQ).", "664079": "Do we need to use tensorflow 2.0 or we are free to use any other framework?\nAlso how we can upload the trained model?", "669677": "Answer is in another thread: https://www.kaggle.com/c/tensorflow2-question-answering/discussion/114727#latest-665122", "676724": "I'm learning machine learning.\nI've been studying for nearly a year, but think of it as a beginner.\nkaggle performed 3 times. (MNIST, Titanic, House Prices)\nI'm still not enough to study, but I want to challenge \"Featured Code Competition\".\nHow should I prepare?\nPlease teach me.", "681505": "Why cannot we use deeppavlov?\n\npython -m deeppavlov install squad_bert\nfrom deeppavlov import build_model, configs\nmodel = build_model(configs.squad.squad, download=True)\nmodel(['DeepPavlov is library for NLP and dialog systems.'], ['What is DeepPavlov?'])\n\nAlso, is this a kernels only competition?", "687255": "How does Kaggle run your notebook (in regards to the input test set) to determine the final winners?\nAnd how must you name your input files/datasets?\nDoes the \"public\" input file have the same name as the \"private\" input file?\n\nLet me give some background and explain my question in detail: (it's a very simple question but it's hard to explain - thus long-winded)\nI'm trying to understand the mechanics of how your notebook (or kernel as I guess they are called) gets run by kaggle to determine \"private leaderboard\" and the ultimate winners of competition.\nIn particular since your notebook will have an input file with a name like:\n'/kaggle/input/tensorflow2-question-answering/simplified-nq-test.jsonl'\nthen how do they run your notebook against the \"private\" test data?\nThe competition rules - description says that they'll run your notebook against a \"private\" data test set.\nSo does this private data set have the same name as the test set they give you?\nDo they simply swap out the 'simplified-nq-test.jsonl' file that they give to each competitor for their own\nsecret data file with the same name 'simplified-nq-test.jsonl'?\nOr is the \"secret\" data file have a different name?\nIf it's different then that means your notebook code must account for that.\n\nThis is really confusing for 1st time competitors as I could find no explanation of this on the competition web page.\nBut when I created a notebook by branching off of a public notebook the above input file name was used within the public notebook.\nThus I can only conclude that is the proper name and the proper subdirectory (data-set) name.\nBut if I'm wrong on this point then my notebook will surely fail when the kaggle-competition committee runs it to determine private leaderboard ranking.\n\nThe whole point of this question is the competitors MUST know how to name the input file within their notebook.", "687307": "I agree that it's not well explained currently, I can't find a source to link to, except for http://blog.kaggle.com/2019/05/17/synchronous-kernels-only-competitions-real-ml-in-real-time/, but this is my understanding:\n- on each submission, the notebook is run twice, first against test set which we can download on the data page, second on unseen private test set\n- file names are the same, only the content is swapped\n- since the notebook is run against both sets, you can be sure that your notebook didn't throw an error when running against private set (if your submission succeeded). But still be sure to make your code perform well on the file you don't see.\n- to avoid probing the test set, you don't see output from your notebook when it's run on private test, and if it's an error, you don't get an error message", "694569": "Are most notebooks using GCP for training? Mine took 4 hours for 8 samples!", "695395": "Did you enable the GPU? Just checking..." }, "source": "meta" }