Is anyone else having difficult downloading train.zip? Would someone who's gotten it successfully mind making it available as a torrent or through some other mechanism?
", "messages": [ { "id": 119945, "postDate": "2016-05-13T21:57:03.233Z", "content": "Hi all,
\n\nWe've split the training dataset into 9 zip files (train_1, train_2, ... train_9) for easier download.
\n\nAlso, train_2 images are now available in Kaggle Scripts.
\n\nEnjoy!!
", "rawMarkdown": "Hi all, \r\n\r\nWe've split the training dataset into 9 zip files (train_1, train_2, ... train_9) for easier download. \r\n\r\nAlso, train_2 images are now available in Kaggle Scripts. \r\n\r\nEnjoy!!", "votes": 3 }, { "id": 119085, "postDate": "2016-05-07T04:51:56.987Z", "content": "I have downloading issues too. I tried it for the second time this night and it aborted after 24 GB. :-( Would be very nice if you could split the data into smaller chunks.
", "rawMarkdown": "I have downloading issues too. I tried it for the second time this night and it aborted after 24 GB. :-( Would be very nice if you could split the data into smaller chunks. ", "votes": 1 }, { "id": 122213, "postDate": "2016-06-02T03:50:27.810Z", "content": "For some reason I am not able to get Chrome option for "Copy as cURL". But found a great Firefox plugin to do the same -
\n\nhttps://addons.mozilla.org/en-US/firefox/addon/cliget/?src=cb-dl-toprated
\n\nHope this helps others
", "rawMarkdown": "For some reason I am not able to get Chrome option for \"Copy as cURL\". But found a great Firefox plugin to do the same - \r\n\r\nhttps://addons.mozilla.org/en-US/firefox/addon/cliget/?src=cb-dl-toprated\r\n\r\nHope this helps others" }, { "id": 119482, "postDate": "2016-05-10T16:43:28.743Z", "content": "Thank you for the hint using the Free-Download-Manager. This solution did it for me even with a low-speed internet connection. ;-)
", "rawMarkdown": "Thank you for the hint using the Free-Download-Manager. This solution did it for me even with a low-speed internet connection. ;-)" }, { "id": 119388, "postDate": "2016-05-09T19:56:31.137Z", "content": "I'd really encourage people to treat downloading the large data set as part of the competition. There are definitely tools out there (ie DownloadThemAll) that will allow you to download large files and to resume the download without restarting it from scratch if the process fails for some reason.
\n\nI know the Genentech competition which closed in January had data files which were around 20 GB. So it's not terribly unreasonable to download something which is twice that size.
", "rawMarkdown": "I'd really encourage people to treat downloading the large data set as part of the competition. There are definitely tools out there (ie DownloadThemAll) that will allow you to download large files and to resume the download without restarting it from scratch if the process fails for some reason.\r\n\r\nI know the Genentech competition which closed in January had data files which were around 20 GB. So it's not terribly unreasonable to download something which is twice that size." }, { "id": 118144, "postDate": "2016-05-02T22:47:41.797Z", "content": "As I explained, that didn't work.
\n\nWhat would probably work very well, is splitting the file into digestible chunks.
", "rawMarkdown": "As I explained, that didn't work. \r\n\r\nWhat would probably work very well, is splitting the file into digestible chunks." }, { "id": 118113, "postDate": "2016-05-02T21:43:23.247Z", "content": "@Amos,
\n\nI'm sorry the downloading experience is frustrating. We do need to stick to browser download because that's how we make sure people have accepted the rules before downloading data. Other users and myself who prefer command line downloads have been using the cookie-wget method that @small yellow duck recommended. Let me know if that doesn't work for you.
\n\n@small yellow duck is our volunteer admin to organize this competition, so please direct the platform requests to me.
", "rawMarkdown": "@Amos, \r\n\r\nI'm sorry the downloading experience is frustrating. We do need to stick to browser download because that's how we make sure people have accepted the rules before downloading data. Other users and myself who prefer command line downloads have been using the [cookie-wget method][1] that @small yellow duck recommended. Let me know if that doesn't work for you. \r\n\r\n@small yellow duck is our volunteer admin to organize this competition, so please direct the platform requests to me. \r\n\r\n [1]: https://www.kaggle.com/forums/f/15/kaggle-forum/t/6604/downloading-data-via-command-line/36206#post36206" }, { "id": 118093, "postDate": "2016-05-02T20:57:14.767Z", "content": "Ducky - the dataset in that other competition was 1/4 the size of just the one train.zip file in this competition. The test file in this competition is about that size, and I had no trouble downloading it.
\n\nCan you simply fix this so its possible for us to download without changing operating systems, browsers, or whatever?
\n\nI don't think its a lot to ask.
", "rawMarkdown": "Ducky - the dataset in that other competition was 1/4 the size of just the one train.zip file in this competition. The test file in this competition is about that size, and I had no trouble downloading it. \r\n\r\nCan you simply fix this so its possible for us to download without changing operating systems, browsers, or whatever? \r\n\r\nI don't think its a lot to ask." }, { "id": 117971, "postDate": "2016-05-02T05:08:21.590Z", "content": "It looks like its actually spamware, and it also seems it may only allow resuming from websites that install their software, but either way its a windows/osx beta, no linux.
\n\nI cannot be the only one who thinks a 36 GB web-only download is a tad on the side of excessive.
", "rawMarkdown": "It looks like its actually spamware, and it also seems it may only allow resuming from websites that install their software, but either way its a windows/osx beta, no linux. \r\n\r\nI cannot be the only one who thinks a 36 GB web-only download is a tad on the side of excessive. " }, { "id": 117970, "postDate": "2016-05-02T04:46:53.920Z", "content": "There is a "Free Download Manager" with a plugin for Chrome which will allow you to pause a download or resume a download which has stopped without having to start from scratch again.
\n\n\n\nedit:\nother Kagglers seem to have had success using the Chrome download manager to acquire large data sets:\nhttps://www.kaggle.com/c/second-annual-data-science-bowl/forums/t/17960/help-how-do-you-guys-download-the-data/101731#post101731
", "rawMarkdown": "There is a \"Free Download Manager\" with a plugin for Chrome which will allow you to pause a download or resume a download which has stopped without having to start from scratch again.\r\n\r\nhttps://chrome.google.com/webstore/detail/free-download-manager-chr/ahmpjcflkgiildlgicmcieglgoilbfdp?hl=en\r\n\r\nedit:\r\nother Kagglers seem to have had success using the Chrome download manager to acquire large data sets:\r\nhttps://www.kaggle.com/c/second-annual-data-science-bowl/forums/t/17960/help-how-do-you-guys-download-the-data/101731#post101731" }, { "id": 117961, "postDate": "2016-05-02T03:39:06.460Z", "content": "I'm using chrome, and I don't see a straightforward way to follow the instructions at that link. Perhaps the training data could be broken-up into smaller chunks?
", "rawMarkdown": "I'm using chrome, and I don't see a straightforward way to follow the instructions at that link. Perhaps the training data could be broken-up into smaller chunks?" }, { "id": 117944, "postDate": "2016-05-01T23:07:16.597Z", "content": "If you're planning to work on your local machine, there's a Firefox plugin called DownThemAll - if a download dies, the plugin will let you resume a download without having to start from scratch.
\n\nIf you're planning to use an AWS instance, you can download the data directly to AWS without first moving it to your local machine - there are some instructions here:\nhttps://www.kaggle.com/forums/f/15/kaggle-forum/t/6604/downloading-data-via-command-line/36206#post36206
", "rawMarkdown": "If you're planning to work on your local machine, there's a Firefox plugin called DownThemAll - if a download dies, the plugin will let you resume a download without having to start from scratch.\r\n\r\nIf you're planning to use an AWS instance, you can download the data directly to AWS without first moving it to your local machine - there are some instructions here:\r\nhttps://www.kaggle.com/forums/f/15/kaggle-forum/t/6604/downloading-data-via-command-line/36206#post36206" }, { "id": 117919, "postDate": "2016-05-01T19:00:35.087Z", "content": "Is anyone else having difficult downloading train.zip? Would someone who's gotten it successfully mind making it available as a torrent or through some other mechanism?
", "rawMarkdown": "Is anyone else having difficult downloading train.zip? Would someone who's gotten it successfully mind making it available as a torrent or through some other mechanism? " }, { "id": 118987, "postDate": "2016-05-06T15:03:46.377Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 119945, "author_name": "Wendy Kan", "author_url": "", "post_date": "2016-05-13T21:57:03.233000", "content": "Hi all,
\n\nWe've split the training dataset into 9 zip files (train_1, train_2, ... train_9) for easier download.
\n\nAlso, train_2 images are now available in Kaggle Scripts.
\n\nEnjoy!!
", "votes": 3, "replies": [] }, { "id": 119085, "author_name": "Laura Fink", "author_url": "", "post_date": "2016-05-07T04:51:56.987000", "content": "I have downloading issues too. I tried it for the second time this night and it aborted after 24 GB. :-( Would be very nice if you could split the data into smaller chunks.
", "votes": 1, "replies": [] }, { "id": 122213, "author_name": "srikara", "author_url": "", "post_date": "2016-06-02T03:50:27.810000", "content": "For some reason I am not able to get Chrome option for "Copy as cURL". But found a great Firefox plugin to do the same -
\n\nhttps://addons.mozilla.org/en-US/firefox/addon/cliget/?src=cb-dl-toprated
\n\nHope this helps others
", "votes": 0, "replies": [] }, { "id": 119482, "author_name": "Laura Fink", "author_url": "", "post_date": "2016-05-10T16:43:28.743000", "content": "Thank you for the hint using the Free-Download-Manager. This solution did it for me even with a low-speed internet connection. ;-)
", "votes": 0, "replies": [] }, { "id": 119388, "author_name": "small yellow duck", "author_url": "", "post_date": "2016-05-09T19:56:31.137000", "content": "I'd really encourage people to treat downloading the large data set as part of the competition. There are definitely tools out there (ie DownloadThemAll) that will allow you to download large files and to resume the download without restarting it from scratch if the process fails for some reason.
\n\nI know the Genentech competition which closed in January had data files which were around 20 GB. So it's not terribly unreasonable to download something which is twice that size.
", "votes": 0, "replies": [] }, { "id": 118144, "author_name": "Amos Elb", "author_url": "", "post_date": "2016-05-02T22:47:41.797000", "content": "As I explained, that didn't work.
\n\nWhat would probably work very well, is splitting the file into digestible chunks.
", "votes": 0, "replies": [] }, { "id": 118113, "author_name": "Wendy Kan", "author_url": "", "post_date": "2016-05-02T21:43:23.247000", "content": "@Amos,
\n\nI'm sorry the downloading experience is frustrating. We do need to stick to browser download because that's how we make sure people have accepted the rules before downloading data. Other users and myself who prefer command line downloads have been using the cookie-wget method that @small yellow duck recommended. Let me know if that doesn't work for you.
\n\n@small yellow duck is our volunteer admin to organize this competition, so please direct the platform requests to me.
", "votes": 0, "replies": [] }, { "id": 118093, "author_name": "Amos Elb", "author_url": "", "post_date": "2016-05-02T20:57:14.767000", "content": "Ducky - the dataset in that other competition was 1/4 the size of just the one train.zip file in this competition. The test file in this competition is about that size, and I had no trouble downloading it.
\n\nCan you simply fix this so its possible for us to download without changing operating systems, browsers, or whatever?
\n\nI don't think its a lot to ask.
", "votes": 0, "replies": [] }, { "id": 117971, "author_name": "Amos Elb", "author_url": "", "post_date": "2016-05-02T05:08:21.590000", "content": "It looks like its actually spamware, and it also seems it may only allow resuming from websites that install their software, but either way its a windows/osx beta, no linux.
\n\nI cannot be the only one who thinks a 36 GB web-only download is a tad on the side of excessive.
", "votes": 0, "replies": [] }, { "id": 117970, "author_name": "small yellow duck", "author_url": "", "post_date": "2016-05-02T04:46:53.920000", "content": "There is a "Free Download Manager" with a plugin for Chrome which will allow you to pause a download or resume a download which has stopped without having to start from scratch again.
\n\n\n\nedit:\nother Kagglers seem to have had success using the Chrome download manager to acquire large data sets:\nhttps://www.kaggle.com/c/second-annual-data-science-bowl/forums/t/17960/help-how-do-you-guys-download-the-data/101731#post101731
", "votes": 0, "replies": [] }, { "id": 117961, "author_name": "Amos Elb", "author_url": "", "post_date": "2016-05-02T03:39:06.460000", "content": "I'm using chrome, and I don't see a straightforward way to follow the instructions at that link. Perhaps the training data could be broken-up into smaller chunks?
", "votes": 0, "replies": [] }, { "id": 117944, "author_name": "small yellow duck", "author_url": "", "post_date": "2016-05-01T23:07:16.597000", "content": "If you're planning to work on your local machine, there's a Firefox plugin called DownThemAll - if a download dies, the plugin will let you resume a download without having to start from scratch.
\n\nIf you're planning to use an AWS instance, you can download the data directly to AWS without first moving it to your local machine - there are some instructions here:\nhttps://www.kaggle.com/forums/f/15/kaggle-forum/t/6604/downloading-data-via-command-line/36206#post36206
", "votes": 0, "replies": [] }, { "id": 118987, "author_name": "", "author_url": "", "post_date": "2016-05-06T15:03:46.377000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "119945": "Hi all, \r\n\r\nWe've split the training dataset into 9 zip files (train_1, train_2, ... train_9) for easier download. \r\n\r\nAlso, train_2 images are now available in Kaggle Scripts. \r\n\r\nEnjoy!!", "119085": "I have downloading issues too. I tried it for the second time this night and it aborted after 24 GB. :-( Would be very nice if you could split the data into smaller chunks. ", "122213": "For some reason I am not able to get Chrome option for \"Copy as cURL\". But found a great Firefox plugin to do the same - \r\n\r\nhttps://addons.mozilla.org/en-US/firefox/addon/cliget/?src=cb-dl-toprated\r\n\r\nHope this helps others", "119482": "Thank you for the hint using the Free-Download-Manager. This solution did it for me even with a low-speed internet connection. ;-)", "119388": "I'd really encourage people to treat downloading the large data set as part of the competition. There are definitely tools out there (ie DownloadThemAll) that will allow you to download large files and to resume the download without restarting it from scratch if the process fails for some reason.\r\n\r\nI know the Genentech competition which closed in January had data files which were around 20 GB. So it's not terribly unreasonable to download something which is twice that size.", "118144": "As I explained, that didn't work. \r\n\r\nWhat would probably work very well, is splitting the file into digestible chunks.", "118113": "@Amos, \r\n\r\nI'm sorry the downloading experience is frustrating. We do need to stick to browser download because that's how we make sure people have accepted the rules before downloading data. Other users and myself who prefer command line downloads have been using the [cookie-wget method][1] that @small yellow duck recommended. Let me know if that doesn't work for you. \r\n\r\n@small yellow duck is our volunteer admin to organize this competition, so please direct the platform requests to me. \r\n\r\n [1]: https://www.kaggle.com/forums/f/15/kaggle-forum/t/6604/downloading-data-via-command-line/36206#post36206", "118093": "Ducky - the dataset in that other competition was 1/4 the size of just the one train.zip file in this competition. The test file in this competition is about that size, and I had no trouble downloading it. \r\n\r\nCan you simply fix this so its possible for us to download without changing operating systems, browsers, or whatever? \r\n\r\nI don't think its a lot to ask.", "117971": "It looks like its actually spamware, and it also seems it may only allow resuming from websites that install their software, but either way its a windows/osx beta, no linux. \r\n\r\nI cannot be the only one who thinks a 36 GB web-only download is a tad on the side of excessive. ", "117970": "There is a \"Free Download Manager\" with a plugin for Chrome which will allow you to pause a download or resume a download which has stopped without having to start from scratch again.\r\n\r\nhttps://chrome.google.com/webstore/detail/free-download-manager-chr/ahmpjcflkgiildlgicmcieglgoilbfdp?hl=en\r\n\r\nedit:\r\nother Kagglers seem to have had success using the Chrome download manager to acquire large data sets:\r\nhttps://www.kaggle.com/c/second-annual-data-science-bowl/forums/t/17960/help-how-do-you-guys-download-the-data/101731#post101731", "117961": "I'm using chrome, and I don't see a straightforward way to follow the instructions at that link. Perhaps the training data could be broken-up into smaller chunks?", "117944": "If you're planning to work on your local machine, there's a Firefox plugin called DownThemAll - if a download dies, the plugin will let you resume a download without having to start from scratch.\r\n\r\nIf you're planning to use an AWS instance, you can download the data directly to AWS without first moving it to your local machine - there are some instructions here:\r\nhttps://www.kaggle.com/forums/f/15/kaggle-forum/t/6604/downloading-data-via-command-line/36206#post36206", "117919": "Is anyone else having difficult downloading train.zip? Would someone who's gotten it successfully mind making it available as a torrent or through some other mechanism? ", "118987": "" } }