![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F2779868%2Fe544b58b1aef2f7ddc623079372b0ba7%2FScreenshot%20from%202019-08-21%2023-06-27.png?generation=1566421647819803&alt=media\")
> @inversion I am sorry to inform you that for 3-5 minutes I published a kernel with a very high score (obviously was a mistake when I tried to share it with my friends due to the kaggle dataset system), please follow the thread to take action. Thank you.
\n\noriginal message:
\n\nF@ck, sorry I shared with my mates and I was slepping, it was when I added my friends… sorry I'm really sorry I didn't realized about that...\nIs the f*ckin kaggle dataset system, I din't mean that
\n\nI feel like those famous persons who leak nude photos on twitter... I was about to cry
\n\nPlease, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray
\n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min I think.
\n\nPS: I'm going to sleep, that's all, I'm really sorry and I canceld it asap
", "messages": [ { "id": 604922, "postDate": "2019-08-21T21:40:28.103Z", "content": "If you shared something publicly then don't remove it. If you remove it then it becomes private sharing and you can get removed from competition.
", "rawMarkdown": "If you shared something publicly then don't remove it. If you remove it then it becomes private sharing and you can get removed from competition.\n", "votes": 11, "replies": [ { "id": 604927, "postDate": "2019-08-21T21:42:20.657Z", "content": "@cpmpml thanks for the comment. I don't want problems, I'll wait for @inversion I don't want to destroy the LB and I'll accept the consequences. Btw, I suppose that rule also applies to the deleted comments in the forum?
", "rawMarkdown": "@cpmpml thanks for the comment. I don't want problems, I'll wait for @inversion I don't want to destroy the LB and I'll accept the consequences. Btw, I suppose that rule also applies to the deleted comments in the forum? ", "votes": 1 }, { "id": 604979, "postDate": "2019-08-21T23:22:35.153Z", "content": "I agree with CPMP. Although high score should not be shared nearby deadline, you cannot cancel your publishment. It's unfair for the people who miss/find your kernel by chance.
", "rawMarkdown": "I agree with CPMP. Although high score should not be shared nearby deadline, you cannot cancel your publishment. It's unfair for the people who miss/find your kernel by chance.", "votes": 3 }, { "id": 605148, "postDate": "2019-08-22T05:26:17.243Z", "content": "Well, @inversion decided otherwise. That's fine with me as long as people play by the rule he enacted.
", "rawMarkdown": "Well, @inversion decided otherwise. That's fine with me as long as people play by the rule he enacted." }, { "id": 605167, "postDate": "2019-08-22T05:58:48.150Z", "content": "@cpmpml Sure, I'll try to stay tuned and do whatever @inversion say. I want a fair competition and that's why I was so nervous and worried.\nSorry.
", "rawMarkdown": "@cpmpml Sure, I'll try to stay tuned and do whatever @inversion say. I want a fair competition and that's why I was so nervous and worried.\nSorry. " }, { "id": 605171, "postDate": "2019-08-22T06:06:26.747Z", "content": "@jesucristo No need to be sorry for me, my rank won't be impacted by this. None of the gold zone will be impacted I think.
", "rawMarkdown": "@jesucristo No need to be sorry for me, my rank won't be impacted by this. None of the gold zone will be impacted I think.", "votes": 1 }, { "id": 605183, "postDate": "2019-08-22T06:32:57.223Z", "content": "@cpmpml lol, who is downvoting everything? even your comments... the guy made my morning, thanks :)
", "rawMarkdown": "@cpmpml lol, who is downvoting everything? even your comments... the guy made my morning, thanks :)", "votes": 1 } ] }, { "id": 604832, "postDate": "2019-08-21T20:17:03.630Z", "content": "Please, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray
\n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min.
\n\n@jiweiliu @khahuras I'm so sorry, I feel so bad, I was about to sleep and my mates called me... I did it as fast as I could
", "rawMarkdown": "Please, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray \n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min.\n\n@jiweiliu @khahuras I'm so sorry, I feel so bad, I was about to sleep and my mates called me... I did it as fast as I could", "votes": 5, "replies": [ { "id": 604838, "postDate": "2019-08-21T20:19:40.773Z", "content": "Will not share, will not use for final sub. I thought it was intentional LB sabotage hence my pissed off comment on the kernel :)
", "rawMarkdown": "Will not share, will not use for final sub. I thought it was intentional LB sabotage hence my pissed off comment on the kernel :)", "votes": 2 }, { "id": 604842, "postDate": "2019-08-21T20:22:03.963Z", "content": "thank you so much @julianmukaj :)
", "rawMarkdown": "thank you so much @julianmukaj :)" }, { "id": 604849, "postDate": "2019-08-21T20:25:36.560Z", "content": "Maybe see if admins can revert those who submitted it, so far just seems to be me and 3 others
", "rawMarkdown": "Maybe see if admins can revert those who submitted it, so far just seems to be me and 3 others", "votes": -1 }, { "id": 604853, "postDate": "2019-08-21T20:30:19.077Z", "content": "There are many teams who could download the submission but haven't upload it yet
", "rawMarkdown": "There are many teams who could download the submission but haven't upload it yet", "votes": 2 }, { "id": 604864, "postDate": "2019-08-21T20:35:54.257Z", "content": "thanks, that's the least you can do, a lot of people have been working really hard for months, I don't know your previous score... but now you are 46th because you were in the right place at the right time. I don't blame you, but obviously I think you don't deserve such score (nothing personal obviously)
", "rawMarkdown": "thanks, that's the least you can do, a lot of people have been working really hard for months, I don't know your previous score... but now you are 46th because you were in the right place at the right time. I don't blame you, but obviously I think you don't deserve such score (nothing personal obviously)", "votes": 2 } ] }, { "id": 604823, "postDate": "2019-08-21T20:11:59.740Z", "content": "> @inversion I am sorry to inform you that for 3-5 minutes I published a kernel with a very high score (obviously was a mistake when I tried to share it with my friends due to the kaggle dataset system), please follow the thread to take action. Thank you.
\n\noriginal message:
\n\nF@ck, sorry I shared with my mates and I was slepping, it was when I added my friends… sorry I'm really sorry I didn't realized about that...\nIs the f*ckin kaggle dataset system, I din't mean that
\n\nI feel like those famous persons who leak nude photos on twitter... I was about to cry
\n\nPlease, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray
\n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min I think.
\n\nPS: I'm going to sleep, that's all, I'm really sorry and I canceld it asap
", "rawMarkdown": "> @inversion I am sorry to inform you that for 3-5 minutes I published a kernel with a very high score (obviously was a mistake when I tried to share it with my friends due to the kaggle dataset system), please follow the thread to take action. Thank you. \n\noriginal message: \n\nF@ck, sorry I shared with my mates and I was slepping, it was when I added my friends… sorry I'm really sorry I didn't realized about that...\nIs the f*ckin kaggle dataset system, I din't mean that\n\nI feel like those famous persons who leak nude photos on twitter... I was about to cry\n\n---\n\nPlease, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray\n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min I think.\n\n---\n\nPS: I'm going to sleep, that's all, I'm really sorry and I canceld it asap", "votes": 2 }, { "id": 604879, "postDate": "2019-08-21T20:54:06.027Z", "content": "@inversion please find a solution quickly.
\n\nI got caught in Talking Data already once. Now, I am going to loose hundred of ranks like many others in the next hours and days.
", "rawMarkdown": "@inversion please find a solution quickly.\n\nI got caught in Talking Data already once. Now, I am going to loose hundred of ranks like many others in the next hours and days.", "votes": 1, "replies": [ { "id": 604885, "postDate": "2019-08-21T20:58:50.647Z", "content": "don't worry, I'm using the LB tracking in orther to see who mysteriously improved in those 3-5 minutes.
", "rawMarkdown": "don't worry, I'm using the LB tracking in orther to see who mysteriously improved in those 3-5 minutes." }, { "id": 604889, "postDate": "2019-08-21T21:00:03.507Z", "content": "The problem is those who downloaded it can submit it in an hour. Or tomorrow. Or 1 min before deadline.
", "rawMarkdown": "The problem is those who downloaded it can submit it in an hour. Or tomorrow. Or 1 min before deadline.", "votes": 7 }, { "id": 604892, "postDate": "2019-08-21T21:02:56.497Z", "content": "yeah... was 3 minutes... I don't kno
", "rawMarkdown": "yeah... was 3 minutes... I don't kno" }, { "id": 604900, "postDate": "2019-08-21T21:09:41.270Z", "content": "I found another case of severe leak, if they change the test data, those persons with good models could do inferences and thats all, my submission leak is useless :) hope they can do it
\n\nin this competition there was a severy leak, and they changed things
\n\n
hope @inversion could do something like this... :)
", "rawMarkdown": "I found another case of severe leak, if they change the test data, those persons with good models could do inferences and thats all, my submission leak is useless :) hope they can do it\n\nin [this competition](https://www.kaggle.com/c/airbus-ship-detection/overview) there was a severy leak, and they changed things\n\n\n\nhope @inversion could do something like this... :)", "votes": -2 }, { "id": 604905, "postDate": "2019-08-21T21:14:45.837Z", "content": "I think it's impossible. I may be wrong, but at the very start someone from organizers wrote thats its impossible, because you cant create new molecules... And if you could, you could get some Nobel prize, not just kaggle medal :D I think the problem of QM9 was disscussed at that moment
", "rawMarkdown": "I think it's impossible. I may be wrong, but at the very start someone from organizers wrote thats its impossible, because you cant create new molecules... And if you could, you could get some Nobel prize, not just kaggle medal :D I think the problem of QM9 was disscussed at that moment", "votes": 1 }, { "id": 604913, "postDate": "2019-08-21T21:24:36.047Z", "content": "@stanislavblinov is not impossible, switch molecules from the public and private, and the order, my submission file will be useless and all the people with real models will be save :)
\n\n> This leaderboard is calculated with approximately 29% of the test data.\nThe final results will be based on the other 71%, so the final standings may be different.
", "rawMarkdown": "@stanislavblinov is not impossible, switch molecules from the public and private, and the order, my submission file will be useless and all the people with real models will be save :)\n\n> This leaderboard is calculated with approximately 29% of the test data.\nThe final results will be based on the other 71%, so the final standings may be different." }, { "id": 604917, "postDate": "2019-08-21T21:31:19.580Z", "content": "how? we know the molecules and atoms we are making predictions on? you can just concat the submission.csv you leaked to the test *set. then you shuffle it to whatever the new order is and resubmit?
", "rawMarkdown": "how? we know the molecules and atoms we are making predictions on? you can just concat the submission.csv you leaked to the test *set. then you shuffle it to whatever the new order is and resubmit? " }, { "id": 605236, "postDate": "2019-08-22T07:24:38.293Z", "content": "@jesucristo I meant about creating totally new test set. \nAnyhow, looks like the solution is found now, so nvm. Let's jest see how it will affect final standings.
", "rawMarkdown": "@jesucristo I meant about creating totally new test set. \nAnyhow, looks like the solution is found now, so nvm. Let's jest see how it will affect final standings." } ] }, { "id": 604835, "postDate": "2019-08-21T20:17:56.217Z", "content": "Don't worry, it's a week away. Everything's going to be OK. !
", "rawMarkdown": "Don't worry, it's a week away. Everything's going to be OK. ! [](https://ru.wikifur.com/w/images/a/ad/Snapshot20121107184927_Droopy-s_good_deed.jpg)", "votes": 1 }, { "id": 604827, "postDate": "2019-08-21T20:13:58.057Z", "content": "Don't be too upset. Let's reach out to admins and see if it can be undone. But what to do with those who have already submitted it ...
", "rawMarkdown": "Don't be too upset. Let's reach out to admins and see if it can be undone. But what to do with those who have already submitted it ...", "votes": 1, "replies": [ { "id": 604851, "postDate": "2019-08-21T20:28:28.857Z", "content": "I wonder how this can be undone...\nI was just downloading other submissions of public kernels to compare and possibly blend with mine, so I have their submission as well and to be honest I don't think I am the only one. And I think there is no way how to undo it as you can not check every submission if it is the same. Obviously, there are people that just copied it, but people could have downloaded it, added a small noise or blend it with their solutions and you would never know.\nI feel like a thief now, but if it was not meant to be public, I am not going to use it. Anyway, I hope I can beat this score with my own model :D
", "rawMarkdown": "I wonder how this can be undone...\nI was just downloading other submissions of public kernels to compare and possibly blend with mine, so I have their submission as well and to be honest I don't think I am the only one. And I think there is no way how to undo it as you can not check every submission if it is the same. Obviously, there are people that just copied it, but people could have downloaded it, added a small noise or blend it with their solutions and you would never know.\nI feel like a thief now, but if it was not meant to be public, I am not going to use it. Anyway, I hope I can beat this score with my own model :D", "votes": 2 }, { "id": 604855, "postDate": "2019-08-21T20:30:50.900Z", "content": "Agree. But how badly I hope to have it undone.
", "rawMarkdown": "Agree. But how badly I hope to have it undone. " }, { "id": 604859, "postDate": "2019-08-21T20:32:10.950Z", "content": "at least the dataset was private... but the submission is all they need for blending the blend of blends of stackings... oh my gosh... hope they can fix it
", "rawMarkdown": "at least the dataset was private... but the submission is all they need for blending the blend of blends of stackings... oh my gosh... hope they can fix it", "votes": -2 }, { "id": 604869, "postDate": "2019-08-21T20:41:08.143Z", "content": "@jiweiliu they are only 4-5 teams, kaggle team can track them for sure (if they really want a fair competition), and I'd like to think that teams at the top will have the courtesy of not use it...
", "rawMarkdown": "@jiweiliu they are only 4-5 teams, kaggle team can track them for sure (if they really want a fair competition), and I'd like to think that teams at the top will have the courtesy of not use it..." }, { "id": 604872, "postDate": "2019-08-21T20:44:36.147Z", "content": "Not only 4-5 teams. Only 4-5 teams submitted it, but many teams downloaded the submission file and they could use it in future
", "rawMarkdown": "Not only 4-5 teams. Only 4-5 teams submitted it, but many teams downloaded the submission file and they could use it in future", "votes": 4 } ] }, { "id": 604899, "postDate": "2019-08-21T21:07:56.593Z", "content": "in this competition there was a severy leak, and they changed things
\n\n
If they change the test data, those teams with good models could do inferences and thats all, my submission leak is useless and everything is fair and clear :) \nhope @inversion could do something like this... :)
\n\nOnce again, I'm so sorry I was about to sleep and now I can't sleep, I feel so bad and furthermore I have to read comments like:
\n\n\n\n\nit's legal because it was public...
\n
omg
", "rawMarkdown": "in [this competition](https://www.kaggle.com/c/airbus-ship-detection/overview) there was a severy leak, and they changed things\n\n\n\nIf they change the test data, those teams with good models could do inferences and thats all, my submission leak is useless and everything is fair and clear :) \nhope @inversion could do something like this... :)\n\nOnce again, I'm so sorry I was about to sleep and now I can't sleep, I feel so bad and furthermore I have to read comments like: \n> it's legal because it was public...\n\nomg", "votes": -2 }, { "id": 604843, "postDate": "2019-08-21T20:22:53.297Z", "content": "Seems like there are just 4 submission. If they won't share anything it can be acceptable as an accident. Shit happens...
", "rawMarkdown": "Seems like there are just 4 submission. If they won't share anything it can be acceptable as an accident. Shit happens...", "votes": -1, "replies": [ { "id": 604846, "postDate": "2019-08-21T20:25:07.523Z", "content": "In my world, even 1 submission makes it completely unfair.
", "rawMarkdown": "In my world, even 1 submission makes it completely unfair. ", "votes": 4 }, { "id": 604847, "postDate": "2019-08-21T20:25:11.813Z", "content": "thanks @silverstone1903 the problem is that one team that got my leak is part of odsai ... you know what I mean? :(
", "rawMarkdown": "thanks @silverstone1903 the problem is that one team that got my leak is part of odsai ... you know what I mean? :(", "votes": -12 }, { "id": 604870, "postDate": "2019-08-21T20:42:19.387Z", "content": "The kernel was public, we submitted it. We didn't know that this kernel was published by mistake. Moreover, many other teams could blend it with their solutions or add little noise, but these teams haven't submitted yet
", "rawMarkdown": "The kernel was public, we submitted it. We didn't know that this kernel was published by mistake. Moreover, many other teams could blend it with their solutions or add little noise, but these teams haven't submitted yet", "votes": 4 }, { "id": 604871, "postDate": "2019-08-21T20:42:19.493Z", "content": "@jesucristo, so what do you mean? By your statement you're showing public disrespect to the community of over 20000 people, and your thoughts are based solely on the behaviour of not more than a couple of individuals. That is very unwise thing to do.
", "rawMarkdown": "@jesucristo, so what do you mean? By your statement you're showing public disrespect to the community of over 20000 people, and your thoughts are based solely on the behaviour of not more than a couple of individuals. That is very unwise thing to do.", "votes": 8 }, { "id": 604875, "postDate": "2019-08-21T20:47:18.960Z", "content": "@vshakhray no no no, I'm part of that amazing community and I think is great place for learning,but that doesn't change the fact that x5 teams (and private exchange) are famous there... And you can find a lot of posts about it\n@canorbal you're right, hope kaggle decide something fair
", "rawMarkdown": "@vshakhray no no no, I'm part of that amazing community and I think is great place for learning,but that doesn't change the fact that *x5 teams* (and private exchange) are famous there... And you can find a lot of posts about it\n@canorbal you're right, hope kaggle decide something fair", "votes": -3 }, { "id": 604881, "postDate": "2019-08-21T20:55:44.527Z", "content": "Only a handful of people downloaded it since it was up for a few mins only. Won't drastically affect medal rankings (I was already in Silver so doesn't change mine), especially not those at the Gold, so I wouldn't be surprised if Kaggle does nothing. I think just toss it up as a mistake and spend this last week trying to improve over it.
", "rawMarkdown": "Only a handful of people downloaded it since it was up for a few mins only. Won't drastically affect medal rankings (I was already in Silver so doesn't change mine), especially not those at the Gold, so I wouldn't be surprised if Kaggle does nothing. I think just toss it up as a mistake and spend this last week trying to improve over it.", "votes": 1 }, { "id": 604883, "postDate": "2019-08-21T20:57:19.873Z", "content": "~Julian you also used it right? that's not fair for others.~
\n\nSorry, I didn't see you mentioned not using it in final submission.
", "rawMarkdown": "~Julian you also used it right? that's not fair for others.~\n\nSorry, I didn't see you mentioned not using it in final submission." }, { "id": 604884, "postDate": "2019-08-21T20:58:20.517Z", "rawMarkdown": "", "votes": -2 }, { "id": 604888, "postDate": "2019-08-21T20:59:39.257Z", "content": "i agree with you on this one. it was there, so why not use it? if i would have been there i would use it too.
", "rawMarkdown": "i agree with you on this one. it was there, so why not use it? if i would have been there i would use it too. " }, { "id": 604890, "postDate": "2019-08-21T21:01:58.770Z", "content": "sure! but it's call fair play, fair competition, some people like me are worried about that :)
", "rawMarkdown": "sure! but it's call **fair play**, fair competition, some people like me are worried about that :)" }, { "id": 604893, "postDate": "2019-08-21T21:04:01.800Z", "content": "@jiweiliu Yes out of respect for the teams mistake I won't use it. My single model just reached -1.95 so leak or not I'd still be in Silver which I am happy with..
", "rawMarkdown": "@jiweiliu Yes out of respect for the teams mistake I won't use it. My single model just reached -1.95 so leak or not I'd still be in Silver which I am happy with.." }, { "id": 604895, "postDate": "2019-08-21T21:04:40.247Z", "content": "its a competition. you obviously made a mistake and i think that is what it is, it happens, but as much as i wouldnt blame you for releasing it, i wouldnt blame them for using it.
", "rawMarkdown": "its a competition. you obviously made a mistake and i think that is what it is, it happens, but as much as i wouldnt blame you for releasing it, i wouldnt blame them for using it.", "votes": 2 }, { "id": 604903, "postDate": "2019-08-21T21:11:20.613Z", "content": "@jiweiliu Yes, you are absolutely right about it. Even one submission is enough but I just wanted to look at the bright side. It could be even worse than this.
", "rawMarkdown": "@jiweiliu Yes, you are absolutely right about it. Even one submission is enough but I just wanted to look at the bright side. It could be even worse than this. " } ] }, { "id": 611452, "postDate": "2019-08-29T09:13:14.847Z", "content": "I think one person used it...
", "rawMarkdown": "I think one person used it..." }, { "id": 609727, "postDate": "2019-08-28T03:35:36.843Z", "content": "life happens... \nunpleasant. not the first time, not the last one either for sure.\nmost importantly there are intentional f..ps and genuine mistakes. I believe there was no bad intention here at all.
", "rawMarkdown": "life happens... \nunpleasant. not the first time, not the last one either for sure.\nmost importantly there are intentional f..ps and genuine mistakes. I believe there was no bad intention here at all." }, { "id": 604936, "postDate": "2019-08-21T22:03:11.043Z", "content": "ok, I'm going to sleep, let's find a solution tomorrow.\nbttw, I'm really sorry (again)
", "rawMarkdown": "ok, I'm going to sleep, let's find a solution tomorrow.\nbttw, I'm really sorry (again)" }, { "id": 604921, "postDate": "2019-08-21T21:40:05.227Z", "content": "Seems that -2.163 is the new baseline of this competition :(
", "rawMarkdown": "Seems that -2.163 is the new baseline of this competition :(", "replies": [ { "id": 605170, "postDate": "2019-08-22T06:04:30.783Z", "content": "Don't think so, it was open for just 3-5 min, I'm completely sure. People didn't shared it and the LB seems stable, I checked the number of views of the kernel and I was really low (50,counting all the time me and my mates refreshed the page) and only 5 forks and 5 submissions using it....\nAnyway, @maxjeblick please accept my apologies, looks like you were affected... Sorry
", "rawMarkdown": "Don't think so, it was open for just 3-5 min, I'm completely sure. People didn't shared it and the LB seems stable, I checked the number of views of the kernel and I was really low (50,counting all the time me and my mates refreshed the page) and only 5 forks and 5 submissions using it....\nAnyway, @maxjeblick please accept my apologies, looks like you were affected... Sorry" } ] }, { "id": 604919, "postDate": "2019-08-21T21:39:36.600Z", "content": "Seems that -2.163 is the new baseline :(
", "rawMarkdown": "Seems that -2.163 is the new baseline :( " }, { "id": 604910, "postDate": "2019-08-21T21:22:48.337Z", "content": "Just what exactly is it in the kernel, does it contain feature/model code beyond the csv files?
", "rawMarkdown": "Just what exactly is it in the kernel, does it contain feature/model code beyond the csv files?", "replies": [ { "id": 604923, "postDate": "2019-08-21T21:40:34.887Z", "content": "nothing, it was just simple stacking... they only got the submission file
", "rawMarkdown": "nothing, it was just simple stacking... they only got the submission file" } ] }, { "id": 604896, "postDate": "2019-08-21T21:04:40.617Z", "content": "I don't know who is downvoting this, instead of make it public and more people will be concerned (included hosts and kaggle) ¿?
", "rawMarkdown": "I don't know who is downvoting this, instead of make it public and more people will be concerned (included hosts and kaggle) ¿? " }, { "id": 604857, "postDate": "2019-08-21T20:31:21.233Z", "content": "how long has it been public ?
", "rawMarkdown": "how long has it been public ?", "replies": [ { "id": 604866, "postDate": "2019-08-21T20:38:06.673Z", "content": "I'm pretty sure 3-5 mins
", "rawMarkdown": "I'm pretty sure 3-5 mins\n" } ] }, { "id": 604831, "postDate": "2019-08-21T20:16:11.607Z", "content": "mistakes happen, this is called \"life\"
", "rawMarkdown": "mistakes happen, this is called \"life\"", "replies": [ { "id": 604837, "postDate": "2019-08-21T20:18:43.680Z", "content": "Sorry @fabiendaniel , you were affected, I feel so bad...
", "rawMarkdown": "Sorry @fabiendaniel , you were affected, I feel so bad..." }, { "id": 604844, "postDate": "2019-08-21T20:23:00.083Z", "content": "no problems, given my current rank, losing 4 places is not a problem. I would have felt worse if at the 3rd place ;)
", "rawMarkdown": "no problems, given my current rank, losing 4 places is not a problem. I would have felt worse if at the 3rd place ;)" } ] }, { "id": 604826, "postDate": "2019-08-21T20:13:47.597Z", "rawMarkdown": "", "isDeleted": true, "replies": [ { "id": 604829, "postDate": "2019-08-21T20:14:50.360Z", "content": "The last time I saw it. it is -2.13 LB
", "rawMarkdown": "The last time I saw it. it is -2.13 LB" }, { "id": 604833, "postDate": "2019-08-21T20:17:23.257Z", "content": "It's too high -2.163
", "rawMarkdown": "It's too high -2.163" } ] } ], "comments": [ { "id": 604922, "author_name": "CPMP", "author_url": "", "post_date": "2019-08-21T21:40:28.103000", "content": "If you shared something publicly then don't remove it. If you remove it then it becomes private sharing and you can get removed from competition.
", "votes": 11, "replies": [ { "id": 604927, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:42:20.657000", "content": "@cpmpml thanks for the comment. I don't want problems, I'll wait for @inversion I don't want to destroy the LB and I'll accept the consequences. Btw, I suppose that rule also applies to the deleted comments in the forum?
", "votes": 1, "replies": [] }, { "id": 604979, "author_name": "toshi_k", "author_url": "", "post_date": "2019-08-21T23:22:35.153000", "content": "I agree with CPMP. Although high score should not be shared nearby deadline, you cannot cancel your publishment. It's unfair for the people who miss/find your kernel by chance.
", "votes": 3, "replies": [] }, { "id": 605148, "author_name": "CPMP", "author_url": "", "post_date": "2019-08-22T05:26:17.243000", "content": "Well, @inversion decided otherwise. That's fine with me as long as people play by the rule he enacted.
", "votes": 0, "replies": [] }, { "id": 605167, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-22T05:58:48.150000", "content": "@cpmpml Sure, I'll try to stay tuned and do whatever @inversion say. I want a fair competition and that's why I was so nervous and worried.\nSorry.
", "votes": 0, "replies": [] }, { "id": 605171, "author_name": "CPMP", "author_url": "", "post_date": "2019-08-22T06:06:26.747000", "content": "@jesucristo No need to be sorry for me, my rank won't be impacted by this. None of the gold zone will be impacted I think.
", "votes": 1, "replies": [] }, { "id": 605183, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-22T06:32:57.223000", "content": "@cpmpml lol, who is downvoting everything? even your comments... the guy made my morning, thanks :)
", "votes": 1, "replies": [] } ] }, { "id": 604832, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:17:03.630000", "content": "Please, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray
\n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min.
\n\n@jiweiliu @khahuras I'm so sorry, I feel so bad, I was about to sleep and my mates called me... I did it as fast as I could
", "votes": 5, "replies": [ { "id": 604838, "author_name": "JM", "author_url": "", "post_date": "2019-08-21T20:19:40.773000", "content": "Will not share, will not use for final sub. I thought it was intentional LB sabotage hence my pissed off comment on the kernel :)
", "votes": 2, "replies": [] }, { "id": 604842, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:22:03.963000", "content": "thank you so much @julianmukaj :)
", "votes": 0, "replies": [] }, { "id": 604849, "author_name": "JM", "author_url": "", "post_date": "2019-08-21T20:25:36.560000", "content": "Maybe see if admins can revert those who submitted it, so far just seems to be me and 3 others
", "votes": -1, "replies": [] }, { "id": 604853, "author_name": "rushakov", "author_url": "", "post_date": "2019-08-21T20:30:19.077000", "content": "There are many teams who could download the submission but haven't upload it yet
", "votes": 2, "replies": [] }, { "id": 604864, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:35:54.257000", "content": "thanks, that's the least you can do, a lot of people have been working really hard for months, I don't know your previous score... but now you are 46th because you were in the right place at the right time. I don't blame you, but obviously I think you don't deserve such score (nothing personal obviously)
", "votes": 2, "replies": [] } ] }, { "id": 604879, "author_name": "eagle4", "author_url": "", "post_date": "2019-08-21T20:54:06.027000", "content": "@inversion please find a solution quickly.
\n\nI got caught in Talking Data already once. Now, I am going to loose hundred of ranks like many others in the next hours and days.
", "votes": 1, "replies": [ { "id": 604885, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:58:50.647000", "content": "don't worry, I'm using the LB tracking in orther to see who mysteriously improved in those 3-5 minutes.
", "votes": 0, "replies": [] }, { "id": 604889, "author_name": "Stanislav Blinov", "author_url": "", "post_date": "2019-08-21T21:00:03.507000", "content": "The problem is those who downloaded it can submit it in an hour. Or tomorrow. Or 1 min before deadline.
", "votes": 7, "replies": [] }, { "id": 604892, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:02:56.497000", "content": "yeah... was 3 minutes... I don't kno
", "votes": 0, "replies": [] }, { "id": 604900, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:09:41.270000", "content": "I found another case of severe leak, if they change the test data, those persons with good models could do inferences and thats all, my submission leak is useless :) hope they can do it
\n\nin this competition there was a severy leak, and they changed things
\n\n
hope @inversion could do something like this... :)
", "votes": -2, "replies": [] }, { "id": 604905, "author_name": "Stanislav Blinov", "author_url": "", "post_date": "2019-08-21T21:14:45.837000", "content": "I think it's impossible. I may be wrong, but at the very start someone from organizers wrote thats its impossible, because you cant create new molecules... And if you could, you could get some Nobel prize, not just kaggle medal :D I think the problem of QM9 was disscussed at that moment
", "votes": 1, "replies": [] }, { "id": 604913, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:24:36.047000", "content": "@stanislavblinov is not impossible, switch molecules from the public and private, and the order, my submission file will be useless and all the people with real models will be save :)
\n\n> This leaderboard is calculated with approximately 29% of the test data.\nThe final results will be based on the other 71%, so the final standings may be different.
", "votes": 0, "replies": [] }, { "id": 604917, "author_name": "Proletheus", "author_url": "", "post_date": "2019-08-21T21:31:19.580000", "content": "how? we know the molecules and atoms we are making predictions on? you can just concat the submission.csv you leaked to the test *set. then you shuffle it to whatever the new order is and resubmit?
", "votes": 0, "replies": [] }, { "id": 605236, "author_name": "Stanislav Blinov", "author_url": "", "post_date": "2019-08-22T07:24:38.293000", "content": "@jesucristo I meant about creating totally new test set. \nAnyhow, looks like the solution is found now, so nvm. Let's jest see how it will affect final standings.
", "votes": 0, "replies": [] } ] }, { "id": 604835, "author_name": "Aleksandr Kosolapov", "author_url": "", "post_date": "2019-08-21T20:17:56.217000", "content": "Don't worry, it's a week away. Everything's going to be OK. !
", "votes": 1, "replies": [] }, { "id": 604827, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2019-08-21T20:13:58.057000", "content": "Don't be too upset. Let's reach out to admins and see if it can be undone. But what to do with those who have already submitted it ...
", "votes": 1, "replies": [ { "id": 604851, "author_name": "somnambWl", "author_url": "", "post_date": "2019-08-21T20:28:28.857000", "content": "I wonder how this can be undone...\nI was just downloading other submissions of public kernels to compare and possibly blend with mine, so I have their submission as well and to be honest I don't think I am the only one. And I think there is no way how to undo it as you can not check every submission if it is the same. Obviously, there are people that just copied it, but people could have downloaded it, added a small noise or blend it with their solutions and you would never know.\nI feel like a thief now, but if it was not meant to be public, I am not going to use it. Anyway, I hope I can beat this score with my own model :D
", "votes": 2, "replies": [] }, { "id": 604855, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2019-08-21T20:30:50.900000", "content": "Agree. But how badly I hope to have it undone.
", "votes": 0, "replies": [] }, { "id": 604859, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:32:10.950000", "content": "at least the dataset was private... but the submission is all they need for blending the blend of blends of stackings... oh my gosh... hope they can fix it
", "votes": -2, "replies": [] }, { "id": 604869, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:41:08.143000", "content": "@jiweiliu they are only 4-5 teams, kaggle team can track them for sure (if they really want a fair competition), and I'd like to think that teams at the top will have the courtesy of not use it...
", "votes": 0, "replies": [] }, { "id": 604872, "author_name": "rushakov", "author_url": "", "post_date": "2019-08-21T20:44:36.147000", "content": "Not only 4-5 teams. Only 4-5 teams submitted it, but many teams downloaded the submission file and they could use it in future
", "votes": 4, "replies": [] } ] }, { "id": 604899, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:07:56.593000", "content": "in this competition there was a severy leak, and they changed things
\n\n
If they change the test data, those teams with good models could do inferences and thats all, my submission leak is useless and everything is fair and clear :) \nhope @inversion could do something like this... :)
\n\nOnce again, I'm so sorry I was about to sleep and now I can't sleep, I feel so bad and furthermore I have to read comments like:
\n\n\n\n\nit's legal because it was public...
\n
omg
", "votes": -2, "replies": [] }, { "id": 604843, "author_name": "silverstone", "author_url": "", "post_date": "2019-08-21T20:22:53.297000", "content": "Seems like there are just 4 submission. If they won't share anything it can be acceptable as an accident. Shit happens...
", "votes": -1, "replies": [ { "id": 604846, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2019-08-21T20:25:07.523000", "content": "In my world, even 1 submission makes it completely unfair.
", "votes": 4, "replies": [] }, { "id": 604847, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:25:11.813000", "content": "thanks @silverstone1903 the problem is that one team that got my leak is part of odsai ... you know what I mean? :(
", "votes": -12, "replies": [] }, { "id": 604870, "author_name": "rushakov", "author_url": "", "post_date": "2019-08-21T20:42:19.387000", "content": "The kernel was public, we submitted it. We didn't know that this kernel was published by mistake. Moreover, many other teams could blend it with their solutions or add little noise, but these teams haven't submitted yet
", "votes": 4, "replies": [] }, { "id": 604871, "author_name": "Vladislav Shakhrai", "author_url": "", "post_date": "2019-08-21T20:42:19.493000", "content": "@jesucristo, so what do you mean? By your statement you're showing public disrespect to the community of over 20000 people, and your thoughts are based solely on the behaviour of not more than a couple of individuals. That is very unwise thing to do.
", "votes": 8, "replies": [] }, { "id": 604875, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:47:18.960000", "content": "@vshakhray no no no, I'm part of that amazing community and I think is great place for learning,but that doesn't change the fact that x5 teams (and private exchange) are famous there... And you can find a lot of posts about it\n@canorbal you're right, hope kaggle decide something fair
", "votes": -3, "replies": [] }, { "id": 604881, "author_name": "JM", "author_url": "", "post_date": "2019-08-21T20:55:44.527000", "content": "Only a handful of people downloaded it since it was up for a few mins only. Won't drastically affect medal rankings (I was already in Silver so doesn't change mine), especially not those at the Gold, so I wouldn't be surprised if Kaggle does nothing. I think just toss it up as a mistake and spend this last week trying to improve over it.
", "votes": 1, "replies": [] }, { "id": 604883, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2019-08-21T20:57:19.873000", "content": "~Julian you also used it right? that's not fair for others.~
\n\nSorry, I didn't see you mentioned not using it in final submission.
", "votes": 0, "replies": [] }, { "id": 604884, "author_name": "rushakov", "author_url": "", "post_date": "2019-08-21T20:58:20.517000", "content": "", "votes": -2, "replies": [] }, { "id": 604888, "author_name": "Proletheus", "author_url": "", "post_date": "2019-08-21T20:59:39.257000", "content": "i agree with you on this one. it was there, so why not use it? if i would have been there i would use it too.
", "votes": 0, "replies": [] }, { "id": 604890, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:01:58.770000", "content": "sure! but it's call fair play, fair competition, some people like me are worried about that :)
", "votes": 0, "replies": [] }, { "id": 604893, "author_name": "JM", "author_url": "", "post_date": "2019-08-21T21:04:01.800000", "content": "@jiweiliu Yes out of respect for the teams mistake I won't use it. My single model just reached -1.95 so leak or not I'd still be in Silver which I am happy with..
", "votes": 0, "replies": [] }, { "id": 604895, "author_name": "Proletheus", "author_url": "", "post_date": "2019-08-21T21:04:40.247000", "content": "its a competition. you obviously made a mistake and i think that is what it is, it happens, but as much as i wouldnt blame you for releasing it, i wouldnt blame them for using it.
", "votes": 2, "replies": [] }, { "id": 604903, "author_name": "silverstone", "author_url": "", "post_date": "2019-08-21T21:11:20.613000", "content": "@jiweiliu Yes, you are absolutely right about it. Even one submission is enough but I just wanted to look at the bright side. It could be even worse than this.
", "votes": 0, "replies": [] } ] }, { "id": 611452, "author_name": "JM", "author_url": "", "post_date": "2019-08-29T09:13:14.847000", "content": "I think one person used it...
", "votes": 0, "replies": [] }, { "id": 609727, "author_name": "Alexey Kotlik", "author_url": "", "post_date": "2019-08-28T03:35:36.843000", "content": "life happens... \nunpleasant. not the first time, not the last one either for sure.\nmost importantly there are intentional f..ps and genuine mistakes. I believe there was no bad intention here at all.
", "votes": 0, "replies": [] }, { "id": 604936, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T22:03:11.043000", "content": "ok, I'm going to sleep, let's find a solution tomorrow.\nbttw, I'm really sorry (again)
", "votes": 0, "replies": [] }, { "id": 604921, "author_name": "Max Jeblick", "author_url": "", "post_date": "2019-08-21T21:40:05.227000", "content": "Seems that -2.163 is the new baseline of this competition :(
", "votes": 0, "replies": [ { "id": 605170, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-22T06:04:30.783000", "content": "Don't think so, it was open for just 3-5 min, I'm completely sure. People didn't shared it and the LB seems stable, I checked the number of views of the kernel and I was really low (50,counting all the time me and my mates refreshed the page) and only 5 forks and 5 submissions using it....\nAnyway, @maxjeblick please accept my apologies, looks like you were affected... Sorry
", "votes": 0, "replies": [] } ] }, { "id": 604919, "author_name": "Max Jeblick", "author_url": "", "post_date": "2019-08-21T21:39:36.600000", "content": "Seems that -2.163 is the new baseline :(
", "votes": 0, "replies": [] }, { "id": 604910, "author_name": "Ren", "author_url": "", "post_date": "2019-08-21T21:22:48.337000", "content": "Just what exactly is it in the kernel, does it contain feature/model code beyond the csv files?
", "votes": 0, "replies": [ { "id": 604923, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:40:34.887000", "content": "nothing, it was just simple stacking... they only got the submission file
", "votes": 0, "replies": [] } ] }, { "id": 604896, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T21:04:40.617000", "content": "I don't know who is downvoting this, instead of make it public and more people will be concerned (included hosts and kaggle) ¿?
", "votes": 0, "replies": [] }, { "id": 604857, "author_name": "FabienDaniel", "author_url": "", "post_date": "2019-08-21T20:31:21.233000", "content": "how long has it been public ?
", "votes": 0, "replies": [ { "id": 604866, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:38:06.673000", "content": "I'm pretty sure 3-5 mins
", "votes": 0, "replies": [] } ] }, { "id": 604831, "author_name": "FabienDaniel", "author_url": "", "post_date": "2019-08-21T20:16:11.607000", "content": "mistakes happen, this is called \"life\"
", "votes": 0, "replies": [ { "id": 604837, "author_name": "Nanashi", "author_url": "", "post_date": "2019-08-21T20:18:43.680000", "content": "Sorry @fabiendaniel , you were affected, I feel so bad...
", "votes": 0, "replies": [] }, { "id": 604844, "author_name": "FabienDaniel", "author_url": "", "post_date": "2019-08-21T20:23:00.083000", "content": "no problems, given my current rank, losing 4 places is not a problem. I would have felt worse if at the 3rd place ;)
", "votes": 0, "replies": [] } ] }, { "id": 604826, "author_name": "", "author_url": "", "post_date": "2019-08-21T20:13:47.597000", "content": "", "votes": 0, "replies": [ { "id": 604829, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2019-08-21T20:14:50.360000", "content": "The last time I saw it. it is -2.13 LB
", "votes": 0, "replies": [] }, { "id": 604833, "author_name": "Sergey Zlobin", "author_url": "", "post_date": "2019-08-21T20:17:23.257000", "content": "It's too high -2.163
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "604922": "If you shared something publicly then don't remove it. If you remove it then it becomes private sharing and you can get removed from competition.\n", "604832": "Please, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray \n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min.\n\n@jiweiliu @khahuras I'm so sorry, I feel so bad, I was about to sleep and my mates called me... I did it as fast as I could", "604823": "> @inversion I am sorry to inform you that for 3-5 minutes I published a kernel with a very high score (obviously was a mistake when I tried to share it with my friends due to the kaggle dataset system), please follow the thread to take action. Thank you. \n\noriginal message: \n\nF@ck, sorry I shared with my mates and I was slepping, it was when I added my friends… sorry I'm really sorry I didn't realized about that...\nIs the f*ckin kaggle dataset system, I din't mean that\n\nI feel like those famous persons who leak nude photos on twitter... I was about to cry\n\n---\n\nPlease, this persons used my leak, do not share nothing for some sad upvotes: @julianmukaj @lesc68 @dukhovnik @canorbal @vshakhray\n\nI'm completely sure they used it and we can check it using @robikscube kernel (LB tracking) My leak was open 3-5 min I think.\n\n---\n\nPS: I'm going to sleep, that's all, I'm really sorry and I canceld it asap", "604879": "@inversion please find a solution quickly.\n\nI got caught in Talking Data already once. Now, I am going to loose hundred of ranks like many others in the next hours and days.", "604835": "Don't worry, it's a week away. Everything's going to be OK. ! [](https://ru.wikifur.com/w/images/a/ad/Snapshot20121107184927_Droopy-s_good_deed.jpg)", "604827": "Don't be too upset. Let's reach out to admins and see if it can be undone. But what to do with those who have already submitted it ...", "604899": "in [this competition](https://www.kaggle.com/c/airbus-ship-detection/overview) there was a severy leak, and they changed things\n\n\n\nIf they change the test data, those teams with good models could do inferences and thats all, my submission leak is useless and everything is fair and clear :) \nhope @inversion could do something like this... :)\n\nOnce again, I'm so sorry I was about to sleep and now I can't sleep, I feel so bad and furthermore I have to read comments like: \n> it's legal because it was public...\n\nomg", "604843": "Seems like there are just 4 submission. If they won't share anything it can be acceptable as an accident. Shit happens...", "611452": "I think one person used it...", "609727": "life happens... \nunpleasant. not the first time, not the last one either for sure.\nmost importantly there are intentional f..ps and genuine mistakes. I believe there was no bad intention here at all.", "604936": "ok, I'm going to sleep, let's find a solution tomorrow.\nbttw, I'm really sorry (again)", "604921": "Seems that -2.163 is the new baseline of this competition :(", "604919": "Seems that -2.163 is the new baseline :( ", "604910": "Just what exactly is it in the kernel, does it contain feature/model code beyond the csv files?", "604896": "I don't know who is downvoting this, instead of make it public and more people will be concerned (included hosts and kaggle) ¿? ", "604857": "how long has it been public ?", "604831": "mistakes happen, this is called \"life\"", "604826": "" } }![](\"https://ru.wikifur.com/w/images/a/ad/Snapshot20121107184927_Droopy-s_good_deed.jpg\"){kind=link}