{ "id": 116049, "title": "Deadline Extension to 11/13!", "url": "/competitions/rsna-intracranial-hemorrhage-detection/discussion/116049", "author_name": "Julia Elliott", "post_date": "2019-11-06T19:27:39.029000", "votes": 28, "comment_count": 43, "views": 0, "content": "

This competition's final submission deadline for Stage 2 has been extended to 11/13 at 11:59pm UTC.

\n\n

As you've noticed, we've experienced 2 separate issues with notebooks and datasets for this competition that are currently our highest priority to resolve.

\n\n

We are attempting to make the full Stage 2 datasets available. In the meantime, we wanted to ensure all participants have the full 6 days you were expecting for Stage 2, so we are extending the deadline.

\n\n

Once the dataset availability issues are fully resolved, we will post another update. We are very sorry for the inconvenience. I promise that we are sharing in your stress and doing everything we can to get things back on track.

", "messages": [ { "id": 667070, "postDate": "2019-11-06T19:27:39.030Z", "content": "

This competition's final submission deadline for Stage 2 has been extended to 11/13 at 11:59pm UTC.

\n\n

As you've noticed, we've experienced 2 separate issues with notebooks and datasets for this competition that are currently our highest priority to resolve.

\n\n

We are attempting to make the full Stage 2 datasets available. In the meantime, we wanted to ensure all participants have the full 6 days you were expecting for Stage 2, so we are extending the deadline.

\n\n

Once the dataset availability issues are fully resolved, we will post another update. We are very sorry for the inconvenience. I promise that we are sharing in your stress and doing everything we can to get things back on track.

", "rawMarkdown": "This competition's final submission deadline for Stage 2 has been extended to **11/13 at 11:59pm UTC**.\n\nAs you've noticed, we've experienced 2 separate issues with notebooks and datasets for this competition that are currently our highest priority to resolve. \n\nWe are attempting to make the full Stage 2 datasets available. In the meantime, we wanted to ensure all participants have the full 6 days you were expecting for Stage 2, so we are extending the deadline.\n\nOnce the dataset availability issues are fully resolved, we will post another update. We are very sorry for the inconvenience. I promise that we are sharing in your stress and doing everything we can to get things back on track.", "votes": 28 }, { "id": 667094, "postDate": "2019-11-06T20:02:00.837Z", "content": "

Always two thoughts when Kaggle deadlines are extended, first thought is \"GREAT, we have more time!\", but also:

\n\n

\"\"

", "rawMarkdown": "Always two thoughts when Kaggle deadlines are extended, first thought is \"GREAT, we have more time!\", but also:\n\n![](https://i.imgflip.com/3fhe7g.jpg)", "votes": 15, "replies": [ { "id": 667438, "postDate": "2019-11-07T08:26:23.367Z", "content": "

Exactly, just like everything else in life, there are always good and bad sides to everything.

", "rawMarkdown": "Exactly, just like everything else in life, there are always good and bad sides to everything.\n", "votes": 2 } ] }, { "id": 667105, "postDate": "2019-11-06T20:32:56.293Z", "content": "

Guys who had enough time to download the whole dataset and train everything on local:\n\"\"

", "rawMarkdown": "Guys who had enough time to download the whole dataset and train everything on local:\n![](https://memepedia.ru/wp-content/uploads/2017/04/evil-toddler.png)", "votes": 11, "replies": [ { "id": 667192, "postDate": "2019-11-06T23:42:08.087Z", "content": "

Lovely child face and I'm longing to have more time to myself as all women on the world.

", "rawMarkdown": "Lovely child face and I'm longing to have more time to myself as all women on the world.", "votes": 1 } ] }, { "id": 668104, "postDate": "2019-11-08T01:42:35.110Z", "content": "

Hi all!

\n\n\n\n

Please let us know if you have questions. As Julia mentioned earlier - we're very sorry for the inconvenience. Thank you for your patience while we worked to find a solution.

", "rawMarkdown": "Hi all!\n\n- Full stage 2 train and test datasets are now available, including within notebooks.\n- If you’re looking to access just the stage 2 test images, you should either download the entire dataset, pull each image systematically via notebooks, or you may find it hosted by another fellow participant elsewhere on the forums.\n- Deadline extension remains 11/13 @ 11:59PM UTC\n\nPlease let us know if you have questions. As Julia mentioned earlier - we're very sorry for the inconvenience. Thank you for your patience while we worked to find a solution.", "votes": 7 }, { "id": 667661, "postDate": "2019-11-07T14:21:52.413Z", "content": "

For those of us who managed to download the datasets for stage 2 (both the train and test ones) shortly after its initial release, should we re-download anything or is it all good?

", "rawMarkdown": "For those of us who managed to download the datasets for stage 2 (both the train and test ones) shortly after its initial release, should we re-download anything or is it all good?", "votes": 5, "replies": [ { "id": 668692, "postDate": "2019-11-08T18:25:15.913Z", "content": "

@juliaelliott @philculliton I have the same question as well. My data with successfully downloaded within 1 day after Stage-1 ended seemed have no problem. Could you please confirm ??

", "rawMarkdown": "@juliaelliott @philculliton I have the same question as well. My data with successfully downloaded within 1 day after Stage-1 ended seemed have no problem. Could you please confirm ??" }, { "id": 668706, "postDate": "2019-11-08T18:50:20.533Z", "content": "

Hi @dsouza and @ratthachat - good question. The stage 2 data has not changed from what was released at the start of stage 2. Just a lot of back end work / magic had to happen to make it more accessible for everyone.

", "rawMarkdown": "Hi @dsouza and @ratthachat - good question. The stage 2 data has not changed from what was released at the start of stage 2. Just a lot of back end work / magic had to happen to make it more accessible for everyone.", "votes": 2 }, { "id": 670540, "postDate": "2019-11-11T15:00:50.540Z", "content": "

Great, thanks a lot!

", "rawMarkdown": "Great, thanks a lot! " } ] }, { "id": 667580, "postDate": "2019-11-07T12:06:01.890Z", "content": "

Great news! Thanks!

\n\n

\"\"

\n\n

@steelrose :)

", "rawMarkdown": "Great news! Thanks! \n\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1788308%2Ff81ac83a8d51ec233a7a8adcb09314dd%2F3fjvtl.jpg?generation=1573128346846677&alt=media)\n\n@steelrose :) \n", "votes": 4 }, { "id": 667545, "postDate": "2019-11-07T11:04:36.107Z", "content": "

IMO the new deadline should be set only AFTER all the issues are resolved.

", "rawMarkdown": "IMO the new deadline should be set only AFTER all the issues are resolved.", "votes": 4, "replies": [ { "id": 667569, "postDate": "2019-11-07T11:53:06.390Z", "content": "

right

", "rawMarkdown": "right", "votes": 2 }, { "id": 667584, "postDate": "2019-11-07T12:17:04.270Z", "content": "

I think so, too. Let's wait for good news.😄

", "rawMarkdown": "I think so, too. Let's wait for good news.😄 " } ] }, { "id": 667576, "postDate": "2019-11-07T12:00:29.007Z", "content": "

\"\"

", "rawMarkdown": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1788308%2Fed84f84025eec1c92a9254fdd82e4bd8%2F3fjvcd.jpg?generation=1573128025405994&alt=media)\n", "votes": 5, "replies": [ { "id": 667583, "postDate": "2019-11-07T12:13:50.067Z", "content": "

Meme thread starts here

", "rawMarkdown": "Meme thread starts here", "votes": 2 } ] }, { "id": 673916, "postDate": "2019-11-15T16:45:06.107Z", "content": "

right

", "rawMarkdown": "right", "votes": 3 }, { "id": 667845, "postDate": "2019-11-07T18:38:09.027Z", "content": "

For clarification - nothing about the Stage 1 train + test set's contents are changing. The labels and images are remaining in-tact. What we are trying to resolve is the issues with these datasets' availability, which has been compromised by the outages/bugs we've experienced. This is still on-going, and we will report back when it has been resolved. In the meantime the extension is intended to subdue panic.

\n\n

Sorry again for the inconvenience; we're getting close to a fix here, but will hold off on announcement until we're certain.

", "rawMarkdown": "For clarification - nothing about the Stage 1 train + test set's contents are changing. The labels and images are remaining in-tact. What we are trying to resolve is the issues with these datasets' availability, which has been compromised by the outages/bugs we've experienced. This is still on-going, and we will report back when it has been resolved. In the meantime the extension is intended to subdue panic. \n\nSorry again for the inconvenience; we're getting close to a fix here, but will hold off on announcement until we're certain.", "votes": 3, "replies": [ { "id": 668043, "postDate": "2019-11-07T22:48:45.793Z", "content": "

Thank you for the generous 48 hour extension and keeping on top of things! I think the continuous communication and transparency throughout the whole process has been professional.

", "rawMarkdown": "Thank you for the generous 48 hour extension and keeping on top of things! I think the continuous communication and transparency throughout the whole process has been professional. ", "votes": 1 } ] }, { "id": 667379, "postDate": "2019-11-07T06:55:58.957Z", "content": "

are you changing dataset pictures or mixing noisy pictures in train set? how do we trust? the train dataset got vanished suddenly without even letting us know! please confirm if you are adding noisy images in train set or not!\nif yes then it will hurt us in stage-2 final leaderboard where those who have downloaded the dataset before for training will get massive advantage. also i would like to know why the train set got vanished suddenly without even letting us know in advance?

", "rawMarkdown": "are you changing dataset pictures or mixing noisy pictures in train set? how do we trust? the train dataset got vanished suddenly without even letting us know! please confirm if you are adding noisy images in train set or not!\nif yes then it will hurt us in stage-2 final leaderboard where those who have downloaded the dataset before for training will get massive advantage. also i would like to know why the train set got vanished suddenly without even letting us know in advance?\n", "votes": 3, "replies": [ { "id": 667439, "postDate": "2019-11-07T08:28:07.653Z", "content": "

You have a good point here. It would be bad and unfair if they suddenly change the data. You should make this thought more visible.

", "rawMarkdown": "You have a good point here. It would be bad and unfair if they suddenly change the data. You should make this thought more visible.", "votes": 2 }, { "id": 667484, "postDate": "2019-11-07T09:35:16.517Z", "content": "

@mobassir I was able to download the stage2 full dataset on the first day in between all the caching and data problems.\nI've did only a small check but it indeed looks like stage1 train + stage1 test = stage2 test.\nI've also looked at some files visually....it doesn't look to be cluttered/changed/noised in any way.

\n\n

So I think we can trust Kaggle on that. Also I'am retraining one of my models on stage 2 training data and the results of the metrics appears to be inline with what I've seen for stage 1.

", "rawMarkdown": "@mobassir I was able to download the stage2 full dataset on the first day in between all the caching and data problems.\nI've did only a small check but it indeed looks like stage1 train + stage1 test = stage2 test.\nI've also looked at some files visually....it doesn't look to be cluttered/changed/noised in any way.\n\nSo I think we can trust Kaggle on that. Also I'am retraining one of my models on stage 2 training data and the results of the metrics appears to be inline with what I've seen for stage 1.", "votes": 2 }, { "id": 667489, "postDate": "2019-11-07T09:40:21.877Z", "content": "

thanks for the confirmation @rsmits

", "rawMarkdown": "thanks for the confirmation @rsmits ", "votes": 1 }, { "id": 667624, "postDate": "2019-11-07T13:15:19.983Z", "content": "

@rsmits you mean stage1train + stage1test = stage2 TRAIN right ?

", "rawMarkdown": "@rsmits you mean stage1train + stage1test = stage2 TRAIN right ? " }, { "id": 667639, "postDate": "2019-11-07T13:43:05.333Z", "content": "

Hi @bdubreu Yes I verified a few labels manually on the .csv files from stage 1 and stage 2. I've not taken the time to verify all of them ;-)

\n\n

From stage 1 train:\nID_63eb1e259\nID_4e6ff6126

\n\n

From stage 1 test:\nID_28fbab7eb\nID_877923b8b

\n\n

Those labels are available in stage_2_train.csv.

", "rawMarkdown": "Hi @bdubreu Yes I verified a few labels manually on the .csv files from stage 1 and stage 2. I've not taken the time to verify all of them ;-)\n\nFrom stage 1 train:\nID_63eb1e259\nID_4e6ff6126\n\nFrom stage 1 test:\nID_28fbab7eb\nID_877923b8b\n\nThose labels are available in stage_2_train.csv." }, { "id": 667724, "postDate": "2019-11-07T15:51:31.197Z", "content": "

I confirm the same. I was confused in the beginning since I saw the data shown as stage 2 train set. But, after unzipping the files I noticed a huge overlap with the stage 1 training set. So looks what they had was stage 1 training and test data together. I think easily double check by looking at the train csv file for stage 2. Good luck everybody.

", "rawMarkdown": "I confirm the same. I was confused in the beginning since I saw the data shown as stage 2 train set. But, after unzipping the files I noticed a huge overlap with the stage 1 training set. So looks what they had was stage 1 training and test data together. I think easily double check by looking at the train csv file for stage 2. Good luck everybody." } ] }, { "id": 667181, "postDate": "2019-11-06T23:06:34.067Z", "content": "

Time a great issue to everyone.

", "rawMarkdown": "Time a great issue to everyone.", "votes": 1 }, { "id": 667175, "postDate": "2019-11-06T22:55:39.450Z", "content": "

Thank you for your decision. So it was great news for me because I was trying to give up👍

", "rawMarkdown": "Thank you for your decision. So it was great news for me because I was trying to give up👍 ", "votes": 1 }, { "id": 668318, "postDate": "2019-11-08T09:10:46.697Z", "content": "

Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DSM files are missing in the folder. Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DCM data are missing in the folder.

\n\n

when will the issue be resolved?

\n\n

\"\"

", "rawMarkdown": "Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DSM files are missing in the folder. Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DCM data are missing in the folder. \n\nwhen will the issue be resolved? \n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F84564%2F39aed108886d89fff179f3919380e65c%2Ferror_kaggle.png?generation=1573204157465836&alt=media)", "votes": 2, "replies": [ { "id": 668362, "postDate": "2019-11-08T10:25:57.907Z", "content": "

Everything works. You just need to fix paths.

", "rawMarkdown": "Everything works. You just need to fix paths.", "votes": 1 }, { "id": 668746, "postDate": "2019-11-08T19:50:07.583Z", "content": "

Hi, @sabarinathan
\nMay be ../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage 2 test/

\n\n

stage 2 test images ---> stage 2 test😄

", "rawMarkdown": "Hi, @sabarinathan \nMay be ../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage 2 test/ \n\nstage 2 test images ---> stage 2 test😄 " }, { "id": 668932, "postDate": "2019-11-09T06:07:02.570Z", "content": "

thanks, I didn't notice the path 👍

", "rawMarkdown": "thanks, I didn't notice the path 👍 " } ] }, { "id": 667437, "postDate": "2019-11-07T08:24:20Z", "content": "

Thank you for the information @juliaelliott.

\n\n

About extension: just like everything else in life, there are always good and bad sides to everything. I don't know whether we should be happy or not with the extension XD

", "rawMarkdown": "Thank you for the information @juliaelliott. \n\nAbout extension: just like everything else in life, there are always good and bad sides to everything. I don't know whether we should be happy or not with the extension XD\n\n", "votes": 2 }, { "id": 679665, "postDate": "2019-11-23T05:56:18.800Z", "content": "

Exactly

", "rawMarkdown": "Exactly" }, { "id": 671684, "postDate": "2019-11-13T03:35:00.947Z", "content": "

When can I choose the final two submissions?

", "rawMarkdown": "When can I choose the final two submissions?", "replies": [ { "id": 671794, "postDate": "2019-11-13T06:47:53.523Z", "content": "

Your team leader needs to select the final two submissions.

", "rawMarkdown": "Your team leader needs to select the final two submissions." } ] }, { "id": 670142, "postDate": "2019-11-11T04:35:45.933Z", "content": "

I've been trying all day to access the images on a kaggle kernel. All I get is is that it's not available. a \"?\" next to the file type. \"This type is not yet supported.\"\nIs it broken again?

", "rawMarkdown": "I've been trying all day to access the images on a kaggle kernel. All I get is is that it's not available. a \"?\" next to the file type. \"This type is not yet supported.\"\nIs it broken again?", "replies": [ { "id": 670236, "postDate": "2019-11-11T07:49:42.843Z", "content": "

Hi, @roushrsh. I also received the same error. Please check your path. It's not just changing 1 to 2.😄

\n\n
\n

'../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_test/'\n '../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_train/'\n '../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_sample_submission.csv'\n '../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_train.csv'

\n
", "rawMarkdown": "Hi, @roushrsh. I also received the same error. Please check your path. It's not just changing 1 to 2.😄 \n> '../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_test/'\n'../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_train/'\n'../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_sample_submission.csv'\n'../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_train.csv'" }, { "id": 671582, "postDate": "2019-11-12T22:57:28.817Z", "content": "

Hehy @mashlyn . I had to add the second rsna-interacranial part which we didn't need to before, even with the data error. Thanks!

", "rawMarkdown": "Hehy @mashlyn . I had to add the second rsna-interacranial part which we didn't need to before, even with the data error. Thanks!", "votes": 1 }, { "id": 671598, "postDate": "2019-11-12T23:47:12.367Z", "content": "

You're welcome😄

", "rawMarkdown": "You're welcome😄 " } ] }, { "id": 669652, "postDate": "2019-11-10T09:45:17.517Z", "content": "

Thank you for the extension. Good news for me.

", "rawMarkdown": "Thank you for the extension. Good news for me." }, { "id": 667695, "postDate": "2019-11-07T15:07:54.453Z", "rawMarkdown": "", "votes": -13, "isDeleted": true, "replies": [ { "id": 667710, "postDate": "2019-11-07T15:36:08.307Z", "content": "

Nice kernel...but....what is the relation between forest fires in brazil and intracranial hemorrhage detection?

", "rawMarkdown": "Nice kernel...but....what is the relation between forest fires in brazil and intracranial hemorrhage detection? ", "votes": 7 } ] }, { "id": 667257, "postDate": "2019-11-07T02:59:10.010Z", "content": "

Thank you so much!!!!!

", "rawMarkdown": "Thank you so much!!!!!", "votes": 1 } ], "comments": [ { "id": 667094, "author_name": "Andrés Miguel Torrubia Sáez", "author_url": "", "post_date": "2019-11-06T20:02:00.837000", "content": "

Always two thoughts when Kaggle deadlines are extended, first thought is \"GREAT, we have more time!\", but also:

\n\n

\"\"

", "votes": 15, "replies": [ { "id": 667438, "author_name": "arctic", "author_url": "", "post_date": "2019-11-07T08:26:23.367000", "content": "

Exactly, just like everything else in life, there are always good and bad sides to everything.

", "votes": 2, "replies": [] } ] }, { "id": 667105, "author_name": "Stanislav Blinov", "author_url": "", "post_date": "2019-11-06T20:32:56.293000", "content": "

Guys who had enough time to download the whole dataset and train everything on local:\n\"\"

", "votes": 11, "replies": [ { "id": 667192, "author_name": "Marília Prata", "author_url": "", "post_date": "2019-11-06T23:42:08.087000", "content": "

Lovely child face and I'm longing to have more time to myself as all women on the world.

", "votes": 1, "replies": [] } ] }, { "id": 668104, "author_name": "Phil Culliton", "author_url": "", "post_date": "2019-11-08T01:42:35.110000", "content": "

Hi all!

\n\n\n\n

Please let us know if you have questions. As Julia mentioned earlier - we're very sorry for the inconvenience. Thank you for your patience while we worked to find a solution.

", "votes": 7, "replies": [] }, { "id": 667661, "author_name": "Daniel Souza", "author_url": "", "post_date": "2019-11-07T14:21:52.413000", "content": "

For those of us who managed to download the datasets for stage 2 (both the train and test ones) shortly after its initial release, should we re-download anything or is it all good?

", "votes": 5, "replies": [ { "id": 668692, "author_name": "Neuron Engineer", "author_url": "", "post_date": "2019-11-08T18:25:15.913000", "content": "

@juliaelliott @philculliton I have the same question as well. My data with successfully downloaded within 1 day after Stage-1 ended seemed have no problem. Could you please confirm ??

", "votes": 0, "replies": [] }, { "id": 668706, "author_name": "Phil Culliton", "author_url": "", "post_date": "2019-11-08T18:50:20.533000", "content": "

Hi @dsouza and @ratthachat - good question. The stage 2 data has not changed from what was released at the start of stage 2. Just a lot of back end work / magic had to happen to make it more accessible for everyone.

", "votes": 2, "replies": [] }, { "id": 670540, "author_name": "Daniel Souza", "author_url": "", "post_date": "2019-11-11T15:00:50.540000", "content": "

Great, thanks a lot!

", "votes": 0, "replies": [] } ] }, { "id": 667580, "author_name": "Mukharbek Organokov", "author_url": "", "post_date": "2019-11-07T12:06:01.890000", "content": "

Great news! Thanks!

\n\n

\"\"

\n\n

@steelrose :)

", "votes": 4, "replies": [] }, { "id": 667545, "author_name": "César Parra Rojas", "author_url": "", "post_date": "2019-11-07T11:04:36.107000", "content": "

IMO the new deadline should be set only AFTER all the issues are resolved.

", "votes": 4, "replies": [ { "id": 667569, "author_name": "Mobassir", "author_url": "", "post_date": "2019-11-07T11:53:06.390000", "content": "

right

", "votes": 2, "replies": [] }, { "id": 667584, "author_name": "Miyabon", "author_url": "", "post_date": "2019-11-07T12:17:04.270000", "content": "

I think so, too. Let's wait for good news.😄

", "votes": 0, "replies": [] } ] }, { "id": 667576, "author_name": "Mukharbek Organokov", "author_url": "", "post_date": "2019-11-07T12:00:29.007000", "content": "

\"\"

", "votes": 5, "replies": [ { "id": 667583, "author_name": "Stanislav Blinov", "author_url": "", "post_date": "2019-11-07T12:13:50.067000", "content": "

Meme thread starts here

", "votes": 2, "replies": [] } ] }, { "id": 673916, "author_name": "Ivan Goncharov", "author_url": "", "post_date": "2019-11-15T16:45:06.107000", "content": "

right

", "votes": 3, "replies": [] }, { "id": 667845, "author_name": "Julia Elliott", "author_url": "", "post_date": "2019-11-07T18:38:09.027000", "content": "

For clarification - nothing about the Stage 1 train + test set's contents are changing. The labels and images are remaining in-tact. What we are trying to resolve is the issues with these datasets' availability, which has been compromised by the outages/bugs we've experienced. This is still on-going, and we will report back when it has been resolved. In the meantime the extension is intended to subdue panic.

\n\n

Sorry again for the inconvenience; we're getting close to a fix here, but will hold off on announcement until we're certain.

", "votes": 3, "replies": [ { "id": 668043, "author_name": "David Tang", "author_url": "", "post_date": "2019-11-07T22:48:45.793000", "content": "

Thank you for the generous 48 hour extension and keeping on top of things! I think the continuous communication and transparency throughout the whole process has been professional.

", "votes": 1, "replies": [] } ] }, { "id": 667379, "author_name": "Mobassir", "author_url": "", "post_date": "2019-11-07T06:55:58.957000", "content": "

are you changing dataset pictures or mixing noisy pictures in train set? how do we trust? the train dataset got vanished suddenly without even letting us know! please confirm if you are adding noisy images in train set or not!\nif yes then it will hurt us in stage-2 final leaderboard where those who have downloaded the dataset before for training will get massive advantage. also i would like to know why the train set got vanished suddenly without even letting us know in advance?

", "votes": 3, "replies": [ { "id": 667439, "author_name": "arctic", "author_url": "", "post_date": "2019-11-07T08:28:07.653000", "content": "

You have a good point here. It would be bad and unfair if they suddenly change the data. You should make this thought more visible.

", "votes": 2, "replies": [] }, { "id": 667484, "author_name": "Robin Smits", "author_url": "", "post_date": "2019-11-07T09:35:16.517000", "content": "

@mobassir I was able to download the stage2 full dataset on the first day in between all the caching and data problems.\nI've did only a small check but it indeed looks like stage1 train + stage1 test = stage2 test.\nI've also looked at some files visually....it doesn't look to be cluttered/changed/noised in any way.

\n\n

So I think we can trust Kaggle on that. Also I'am retraining one of my models on stage 2 training data and the results of the metrics appears to be inline with what I've seen for stage 1.

", "votes": 2, "replies": [] }, { "id": 667489, "author_name": "Mobassir", "author_url": "", "post_date": "2019-11-07T09:40:21.877000", "content": "

thanks for the confirmation @rsmits

", "votes": 1, "replies": [] }, { "id": 667624, "author_name": "Benjamin Dubreu", "author_url": "", "post_date": "2019-11-07T13:15:19.983000", "content": "

@rsmits you mean stage1train + stage1test = stage2 TRAIN right ?

", "votes": 0, "replies": [] }, { "id": 667639, "author_name": "Robin Smits", "author_url": "", "post_date": "2019-11-07T13:43:05.333000", "content": "

Hi @bdubreu Yes I verified a few labels manually on the .csv files from stage 1 and stage 2. I've not taken the time to verify all of them ;-)

\n\n

From stage 1 train:\nID_63eb1e259\nID_4e6ff6126

\n\n

From stage 1 test:\nID_28fbab7eb\nID_877923b8b

\n\n

Those labels are available in stage_2_train.csv.

", "votes": 0, "replies": [] }, { "id": 667724, "author_name": "Hilal Shaath", "author_url": "", "post_date": "2019-11-07T15:51:31.197000", "content": "

I confirm the same. I was confused in the beginning since I saw the data shown as stage 2 train set. But, after unzipping the files I noticed a huge overlap with the stage 1 training set. So looks what they had was stage 1 training and test data together. I think easily double check by looking at the train csv file for stage 2. Good luck everybody.

", "votes": 0, "replies": [] } ] }, { "id": 667181, "author_name": "Marília Prata", "author_url": "", "post_date": "2019-11-06T23:06:34.067000", "content": "

Time a great issue to everyone.

", "votes": 1, "replies": [] }, { "id": 667175, "author_name": "Miyabon", "author_url": "", "post_date": "2019-11-06T22:55:39.450000", "content": "

Thank you for your decision. So it was great news for me because I was trying to give up👍

", "votes": 1, "replies": [] }, { "id": 668318, "author_name": "sabari nathan", "author_url": "", "post_date": "2019-11-08T09:10:46.697000", "content": "

Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DSM files are missing in the folder. Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DCM data are missing in the folder.

\n\n

when will the issue be resolved?

\n\n

\"\"

", "votes": 2, "replies": [ { "id": 668362, "author_name": "Stanislav Blinov", "author_url": "", "post_date": "2019-11-08T10:25:57.907000", "content": "

Everything works. You just need to fix paths.

", "votes": 1, "replies": [] }, { "id": 668746, "author_name": "Miyabon", "author_url": "", "post_date": "2019-11-08T19:50:07.583000", "content": "

Hi, @sabarinathan
\nMay be ../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage 2 test/

\n\n

stage 2 test images ---> stage 2 test😄

", "votes": 0, "replies": [] }, { "id": 668932, "author_name": "sabari nathan", "author_url": "", "post_date": "2019-11-09T06:07:02.570000", "content": "

thanks, I didn't notice the path 👍

", "votes": 0, "replies": [] } ] }, { "id": 667437, "author_name": "arctic", "author_url": "", "post_date": "2019-11-07T08:24:20", "content": "

Thank you for the information @juliaelliott.

\n\n

About extension: just like everything else in life, there are always good and bad sides to everything. I don't know whether we should be happy or not with the extension XD

", "votes": 2, "replies": [] }, { "id": 679665, "author_name": "Bhuvan Chadha", "author_url": "", "post_date": "2019-11-23T05:56:18.800000", "content": "

Exactly

", "votes": 0, "replies": [] }, { "id": 671684, "author_name": "qzw", "author_url": "", "post_date": "2019-11-13T03:35:00.947000", "content": "

When can I choose the final two submissions?

", "votes": 0, "replies": [ { "id": 671794, "author_name": "Alison Davey", "author_url": "", "post_date": "2019-11-13T06:47:53.523000", "content": "

Your team leader needs to select the final two submissions.

", "votes": 0, "replies": [] } ] }, { "id": 670142, "author_name": "roush", "author_url": "", "post_date": "2019-11-11T04:35:45.933000", "content": "

I've been trying all day to access the images on a kaggle kernel. All I get is is that it's not available. a \"?\" next to the file type. \"This type is not yet supported.\"\nIs it broken again?

", "votes": 0, "replies": [ { "id": 670236, "author_name": "Miyabon", "author_url": "", "post_date": "2019-11-11T07:49:42.843000", "content": "

Hi, @roushrsh. I also received the same error. Please check your path. It's not just changing 1 to 2.😄

\n\n
\n

'../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_test/'\n '../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_train/'\n '../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_sample_submission.csv'\n '../input/rsna-intracranial-hemorrhage-detection/rsna-intracranial-hemorrhage-detection/stage_2_train.csv'

\n
", "votes": 0, "replies": [] }, { "id": 671582, "author_name": "roush", "author_url": "", "post_date": "2019-11-12T22:57:28.817000", "content": "

Hehy @mashlyn . I had to add the second rsna-interacranial part which we didn't need to before, even with the data error. Thanks!

", "votes": 1, "replies": [] }, { "id": 671598, "author_name": "Miyabon", "author_url": "", "post_date": "2019-11-12T23:47:12.367000", "content": "

You're welcome😄

", "votes": 0, "replies": [] } ] }, { "id": 669652, "author_name": "Naruhiko Nakanishi", "author_url": "", "post_date": "2019-11-10T09:45:17.517000", "content": "

Thank you for the extension. Good news for me.

", "votes": 0, "replies": [] }, { "id": 667695, "author_name": "", "author_url": "", "post_date": "2019-11-07T15:07:54.453000", "content": "", "votes": -13, "replies": [ { "id": 667710, "author_name": "Robin Smits", "author_url": "", "post_date": "2019-11-07T15:36:08.307000", "content": "

Nice kernel...but....what is the relation between forest fires in brazil and intracranial hemorrhage detection?

", "votes": 7, "replies": [] } ] }, { "id": 667257, "author_name": "Hilal Shaath", "author_url": "", "post_date": "2019-11-07T02:59:10.010000", "content": "

Thank you so much!!!!!

", "votes": 1, "replies": [] } ], "raw_markdown_by_id": { "667070": "This competition's final submission deadline for Stage 2 has been extended to **11/13 at 11:59pm UTC**.\n\nAs you've noticed, we've experienced 2 separate issues with notebooks and datasets for this competition that are currently our highest priority to resolve. \n\nWe are attempting to make the full Stage 2 datasets available. In the meantime, we wanted to ensure all participants have the full 6 days you were expecting for Stage 2, so we are extending the deadline.\n\nOnce the dataset availability issues are fully resolved, we will post another update. We are very sorry for the inconvenience. I promise that we are sharing in your stress and doing everything we can to get things back on track.", "667094": "Always two thoughts when Kaggle deadlines are extended, first thought is \"GREAT, we have more time!\", but also:\n\n![](https://i.imgflip.com/3fhe7g.jpg)", "667105": "Guys who had enough time to download the whole dataset and train everything on local:\n![](https://memepedia.ru/wp-content/uploads/2017/04/evil-toddler.png)", "668104": "Hi all!\n\n- Full stage 2 train and test datasets are now available, including within notebooks.\n- If you’re looking to access just the stage 2 test images, you should either download the entire dataset, pull each image systematically via notebooks, or you may find it hosted by another fellow participant elsewhere on the forums.\n- Deadline extension remains 11/13 @ 11:59PM UTC\n\nPlease let us know if you have questions. As Julia mentioned earlier - we're very sorry for the inconvenience. Thank you for your patience while we worked to find a solution.", "667661": "For those of us who managed to download the datasets for stage 2 (both the train and test ones) shortly after its initial release, should we re-download anything or is it all good?", "667580": "Great news! Thanks! \n\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1788308%2Ff81ac83a8d51ec233a7a8adcb09314dd%2F3fjvtl.jpg?generation=1573128346846677&alt=media)\n\n@steelrose :) \n", "667545": "IMO the new deadline should be set only AFTER all the issues are resolved.", "667576": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F1788308%2Fed84f84025eec1c92a9254fdd82e4bd8%2F3fjvcd.jpg?generation=1573128025405994&alt=media)\n", "673916": "right", "667845": "For clarification - nothing about the Stage 1 train + test set's contents are changing. The labels and images are remaining in-tact. What we are trying to resolve is the issues with these datasets' availability, which has been compromised by the outages/bugs we've experienced. This is still on-going, and we will report back when it has been resolved. In the meantime the extension is intended to subdue panic. \n\nSorry again for the inconvenience; we're getting close to a fix here, but will hold off on announcement until we're certain.", "667379": "are you changing dataset pictures or mixing noisy pictures in train set? how do we trust? the train dataset got vanished suddenly without even letting us know! please confirm if you are adding noisy images in train set or not!\nif yes then it will hurt us in stage-2 final leaderboard where those who have downloaded the dataset before for training will get massive advantage. also i would like to know why the train set got vanished suddenly without even letting us know in advance?\n", "667181": "Time a great issue to everyone.", "667175": "Thank you for your decision. So it was great news for me because I was trying to give up👍 ", "668318": "Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DSM files are missing in the folder. Hi Team, Still I am not able to access the stage 2 train and test images from notebooks. I can read the CSV files, but I am getting DCM data are missing in the folder. \n\nwhen will the issue be resolved? \n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F84564%2F39aed108886d89fff179f3919380e65c%2Ferror_kaggle.png?generation=1573204157465836&alt=media)", "667437": "Thank you for the information @juliaelliott. \n\nAbout extension: just like everything else in life, there are always good and bad sides to everything. I don't know whether we should be happy or not with the extension XD\n\n", "679665": "Exactly", "671684": "When can I choose the final two submissions?", "670142": "I've been trying all day to access the images on a kaggle kernel. All I get is is that it's not available. a \"?\" next to the file type. \"This type is not yet supported.\"\nIs it broken again?", "669652": "Thank you for the extension. Good news for me.", "667695": "", "667257": "Thank you so much!!!!!" } }