![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F933480%2F1b76cf93338e451e84ee929fc7ecbefd%2FScreen%20Shot%202020-08-19%20at%204.58.03%20PM.png?generation=1597874298812540&alt=media\")
I know a lot of work went into the finalization of the LB, so I'd like to thank Kaggle for that.
\nIt seems I've been mistakenly removed from the LB. I joined this competition very late, way after the whole yolo fiasco, so I never bothered to use yolo and only submitted effdet and fasterrcnn models. I never use any external data. Further I'm too \"old\" to cheat, so I figured it must be some sort of mistake. Currently, my subs look like this:
\n
Since there is at least one valid submission, I figured I should minimally be represented on the LB. That I'm not there at all is alarming, but I'm certain it's a small mistake the Kaggle staff can clear upon inspection.
\nWhat is much more alarming to me:
\nThis is my fasterrcnn model: https://www.kaggle.com/authman/faster-rcnn-inference1?scriptVersionId=40045711 I submitted this model before the original end date of the competition and it ran successfully with a silver medal score reported. Training the model took about 18 hours. I saved the folds into a kaggle dataset to do inference, per the competition rules.
\nSince I am also participating in the melanoma competition as well as the fibrosis competition, in order to clear up some private dataset space, I deleted the folds used to train this model. The deletion occurred days after the conclusion of this competition. Never in my wildest dreams did I think that the actual models would be re-ran again. Up until yesterday this submission had a score, but today it does not. Is this due to the dataset being deleted—or some other issue?
\nAny clarity or direction that could be provided here would be greatly appreciated. Thank you for your attention to this! @philculliton @maggiemd
", "messages": [ { "id": 978043, "postDate": "2020-08-19T22:06:39.180Z", "content": "I know a lot of work went into the finalization of the LB, so I'd like to thank Kaggle for that.
\nIt seems I've been mistakenly removed from the LB. I joined this competition very late, way after the whole yolo fiasco, so I never bothered to use yolo and only submitted effdet and fasterrcnn models. I never use any external data. Further I'm too \"old\" to cheat, so I figured it must be some sort of mistake. Currently, my subs look like this:
\n
Since there is at least one valid submission, I figured I should minimally be represented on the LB. That I'm not there at all is alarming, but I'm certain it's a small mistake the Kaggle staff can clear upon inspection.
\nWhat is much more alarming to me:
\nThis is my fasterrcnn model: https://www.kaggle.com/authman/faster-rcnn-inference1?scriptVersionId=40045711 I submitted this model before the original end date of the competition and it ran successfully with a silver medal score reported. Training the model took about 18 hours. I saved the folds into a kaggle dataset to do inference, per the competition rules.
\nSince I am also participating in the melanoma competition as well as the fibrosis competition, in order to clear up some private dataset space, I deleted the folds used to train this model. The deletion occurred days after the conclusion of this competition. Never in my wildest dreams did I think that the actual models would be re-ran again. Up until yesterday this submission had a score, but today it does not. Is this due to the dataset being deleted—or some other issue?
\nAny clarity or direction that could be provided here would be greatly appreciated. Thank you for your attention to this! @philculliton @maggiemd
", "rawMarkdown": "I know a lot of work went into the finalization of the LB, so I'd like to thank Kaggle for that.\n\nIt seems I've been mistakenly removed from the LB. I joined this competition very late, way after the whole yolo fiasco, so I never bothered to use yolo and only submitted effdet and fasterrcnn models. I never use any external data. Further I'm too \"old\" to cheat, so I figured it must be some sort of mistake. Currently, my subs look like this:\n\n\n\nSince there is at least one valid submission, I figured I should minimally be represented on the LB. That I'm not there at all is alarming, but I'm certain it's a small mistake the Kaggle staff can clear upon inspection.\n\nWhat is **much more alarming** to me:\n\nThis is my fasterrcnn model: https://www.kaggle.com/authman/faster-rcnn-inference1?scriptVersionId=40045711 I submitted this model before the original end date of the competition and it ran successfully with a silver medal score reported. Training the model took about 18 hours. I saved the folds into a kaggle dataset to do inference, per the competition rules.\n\nSince I am also participating in the melanoma competition as well as the fibrosis competition, in order to clear up some private dataset space, I deleted the folds used to train this model. The deletion occurred **days after** the conclusion of this competition. Never in my wildest dreams did I think that the actual models would be re-ran again. Up until yesterday this submission had a score, but today it does not. Is this due to the dataset being deleted—or some other issue?\n\nAny clarity or direction that could be provided here would be greatly appreciated. Thank you for your attention to this! @philculliton @maggiemd", "votes": 11 }, { "id": 978105, "postDate": "2020-08-20T00:25:58.700Z", "content": "I think your best bet is to contact the Kaggle compliance team here: https://www.kaggle.com/compliance
", "rawMarkdown": "I think your best bet is to contact the Kaggle compliance team here: https://www.kaggle.com/compliance", "votes": 3, "replies": [ { "id": 978118, "postDate": "2020-08-20T00:57:54.570Z", "content": "Solid suggestion. Just submitted!
", "rawMarkdown": "Solid suggestion. Just submitted!" } ] }, { "id": 978214, "postDate": "2020-08-20T03:08:36.080Z", "content": "same happened to our team. if you are able to do any progress please do tell us.
", "rawMarkdown": "same happened to our team. if you are able to do any progress please do tell us.", "votes": 1 }, { "id": 978174, "postDate": "2020-08-20T02:21:04.237Z", "content": "I met the same problem.🤕 here
", "rawMarkdown": "I met the same problem.🤕 [here](https://www.kaggle.com/c/global-wheat-detection/discussion/175956)", "votes": 2 }, { "id": 978109, "postDate": "2020-08-20T00:32:11.860Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 978105, "author_name": "Matt", "author_url": "", "post_date": "2020-08-20T00:25:58.700000", "content": "I think your best bet is to contact the Kaggle compliance team here: https://www.kaggle.com/compliance
", "votes": 3, "replies": [ { "id": 978118, "author_name": "عثمان", "author_url": "", "post_date": "2020-08-20T00:57:54.570000", "content": "Solid suggestion. Just submitted!
", "votes": 0, "replies": [] } ] }, { "id": 978214, "author_name": "master_roshi", "author_url": "", "post_date": "2020-08-20T03:08:36.080000", "content": "same happened to our team. if you are able to do any progress please do tell us.
", "votes": 1, "replies": [] }, { "id": 978174, "author_name": "Dapeng", "author_url": "", "post_date": "2020-08-20T02:21:04.237000", "content": "I met the same problem.🤕 here
", "votes": 2, "replies": [] }, { "id": 978109, "author_name": "", "author_url": "", "post_date": "2020-08-20T00:32:11.860000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "978043": "I know a lot of work went into the finalization of the LB, so I'd like to thank Kaggle for that.\n\nIt seems I've been mistakenly removed from the LB. I joined this competition very late, way after the whole yolo fiasco, so I never bothered to use yolo and only submitted effdet and fasterrcnn models. I never use any external data. Further I'm too \"old\" to cheat, so I figured it must be some sort of mistake. Currently, my subs look like this:\n\n\n\nSince there is at least one valid submission, I figured I should minimally be represented on the LB. That I'm not there at all is alarming, but I'm certain it's a small mistake the Kaggle staff can clear upon inspection.\n\nWhat is **much more alarming** to me:\n\nThis is my fasterrcnn model: https://www.kaggle.com/authman/faster-rcnn-inference1?scriptVersionId=40045711 I submitted this model before the original end date of the competition and it ran successfully with a silver medal score reported. Training the model took about 18 hours. I saved the folds into a kaggle dataset to do inference, per the competition rules.\n\nSince I am also participating in the melanoma competition as well as the fibrosis competition, in order to clear up some private dataset space, I deleted the folds used to train this model. The deletion occurred **days after** the conclusion of this competition. Never in my wildest dreams did I think that the actual models would be re-ran again. Up until yesterday this submission had a score, but today it does not. Is this due to the dataset being deleted—or some other issue?\n\nAny clarity or direction that could be provided here would be greatly appreciated. Thank you for your attention to this! @philculliton @maggiemd", "978105": "I think your best bet is to contact the Kaggle compliance team here: https://www.kaggle.com/compliance", "978214": "same happened to our team. if you are able to do any progress please do tell us.", "978174": "I met the same problem.🤕 [here](https://www.kaggle.com/c/global-wheat-detection/discussion/175956)", "978109": "" } }