Here is my last submission code yolov4 + effdicient d7.I had checked many times to insure my code following the rules. So I have no idea which library I used is prohibited. Hope kaggle team or someone can help me. Thanks!!
", "messages": [ { "id": 977488, "postDate": "2020-08-19T14:06:25.943Z", "content": "Here is my last submission code yolov4 + effdicient d7.I had checked many times to insure my code following the rules. So I have no idea which library I used is prohibited. Hope kaggle team or someone can help me. Thanks!!
", "rawMarkdown": "Here is my last submission code [yolov4 + effdicient d7](https://www.kaggle.com/marcowu/ap2-0effd-yolov4-pseudo-labeling-oof-evalu?scriptVersionId=39989367).I had checked many times to insure my code following the rules. So I have no idea which library I used is prohibited. Hope kaggle team or someone can help me. Thanks!! ", "votes": 9 }, { "id": 977723, "postDate": "2020-08-19T17:03:30.423Z", "content": "I paste here my answer from the same topic.
\nIt seems like one of these libs has a GPL-3 license which is prohibited for this competition:
\nThank you, I will check it again.
", "rawMarkdown": "Thank you, I will check it again." } ] }, { "id": 977655, "postDate": "2020-08-19T16:15:02.830Z", "content": "I'm sorry this happened to you. I don't see anything that could be illegal, though it is quite similar to NVNN's yolov5 notebook. I have a very similar notebook (yolov4 based on NVNN's pseudolabelling notebook) that was not invalidated; perhaps this will help understand. Perhaps you could also email someone on the Kaggle team.
\nhttps://www.kaggle.com/stanleyjzheng/yolov4pl-oof
", "rawMarkdown": "I'm sorry this happened to you. I don't see anything that could be illegal, though it is quite similar to NVNN's yolov5 notebook. I have a very similar notebook (yolov4 based on NVNN's pseudolabelling notebook) that was not invalidated; perhaps this will help understand. Perhaps you could also email someone on the Kaggle team. \n\nhttps://www.kaggle.com/stanleyjzheng/yolov4pl-oof", "votes": 1, "replies": [ { "id": 977665, "postDate": "2020-08-19T16:21:57.173Z", "rawMarkdown": "", "isDeleted": true }, { "id": 977668, "postDate": "2020-08-19T16:26:09.970Z", "content": "Thank you very much!! I already send an email to Phil Culliton. Hope they will give me a reply in 8 hours.
", "rawMarkdown": "Thank you very much!! I already send an email to Phil Culliton. Hope they will give me a reply in 8 hours.", "votes": 1 }, { "id": 977683, "postDate": "2020-08-19T16:36:36.297Z", "content": "@stanleyjzheng My ground truth's csv_file was changed with my yolov5 distribution, so I named it as yolov5trainF5_onlyfix.csv. Perhaps it is one reason cause me error,but I think this should not be illegal.
\nThank you very much!!
", "rawMarkdown": "@stanleyjzheng My ground truth's csv_file was changed with my yolov5 distribution, so I named it as yolov5trainF5_onlyfix.csv. Perhaps it is one reason cause me error,but I think this should not be illegal.\n\nThank you very much!!", "votes": 1 }, { "id": 977728, "postDate": "2020-08-19T17:10:47.370Z", "content": "I agree it definitely shouldn't, but it's possible they're just scanning for yolov5 in every notebook and invalidating them that way (more likely it's way more advanced or done by hand though, we'll see). Best of luck with your email.
", "rawMarkdown": "I agree it definitely shouldn't, but it's possible they're just scanning for yolov5 in every notebook and invalidating them that way (more likely it's way more advanced or done by hand though, we'll see). Best of luck with your email.", "votes": 1 }, { "id": 977754, "postDate": "2020-08-19T17:30:34.373Z", "content": "Thanks!! Wish you can get great private score!
", "rawMarkdown": "Thanks!! Wish you can get great private score!", "votes": 1 } ] }, { "id": 977628, "postDate": "2020-08-19T15:51:05.543Z", "content": "@marcowu - I dont see any prohibited libraries as well. My only guess is it would have taken more than 6 hrs for the execution to complete.
", "rawMarkdown": "@marcowu - I dont see any prohibited libraries as well. My only guess is it would have taken more than 6 hrs for the execution to complete. ", "votes": 1, "replies": [ { "id": 977639, "postDate": "2020-08-19T15:56:35.643Z", "content": "My one submission was only running inference, not pseudo labeling.
", "rawMarkdown": "My one submission was only running inference, not pseudo labeling.", "votes": 1 } ] }, { "id": 977600, "postDate": "2020-08-19T15:34:20.540Z", "content": "@philculliton Why this notebook even could get score,but I can not?
\n", "rawMarkdown": "@philculliton Why this notebook even could get score,but I can not?\n\nhttps://www.kaggle.com/naimur978/wheat-head-detection-yolov5-pseudo-labelling-oof?scriptVersionId=38595668", "votes": 1 }, { "id": 977690, "postDate": "2020-08-19T16:40:41.733Z", "content": "Maybe because you are using RAdam in your dataset (not sure). I actually got more boost using RAdam but didn't use it as it was nowhere mentions in the external data thread. I even asked as a discussion topic but ended with zero response.
\nThis is just my speculation, I might be wrong here.
Thanks your reply!! My radam.py license is AP 2.0 in here. It is very disappoint that no kaggle team reply in your discussion. Because my one submission was only running inference, did not use radam. If it truly caused me out LB,I will hard to accept.
", "rawMarkdown": "Thanks your reply!! My radam.py license is AP 2.0 in [here](https://github.com/LiyuanLucasLiu/RAdam/blob/master/LICENSE). It is very disappoint that no kaggle team reply in your discussion. Because my one submission was only running inference, did not use radam. If it truly caused me out LB,I will hard to accept. ", "votes": 1 }, { "id": 977820, "postDate": "2020-08-19T18:09:28.213Z", "content": "I see that you did not use it in your inference code. If this is the reason being invalidated then I think the Kaggle team needs to reconsider your submission.
\nIt's best to reach out to the Kaggle team and clarify it. Hope they will restore your submission. Best of luck.
I paste here my answer from the same topic.
\nIt seems like one of these libs has a GPL-3 license which is prohibited for this competition:
\nThank you, I will check it again.
", "votes": 0, "replies": [] } ] }, { "id": 977655, "author_name": "Stanley Zheng", "author_url": "", "post_date": "2020-08-19T16:15:02.830000", "content": "I'm sorry this happened to you. I don't see anything that could be illegal, though it is quite similar to NVNN's yolov5 notebook. I have a very similar notebook (yolov4 based on NVNN's pseudolabelling notebook) that was not invalidated; perhaps this will help understand. Perhaps you could also email someone on the Kaggle team.
\nhttps://www.kaggle.com/stanleyjzheng/yolov4pl-oof
", "votes": 1, "replies": [ { "id": 977665, "author_name": "", "author_url": "", "post_date": "2020-08-19T16:21:57.173000", "content": "", "votes": 0, "replies": [] }, { "id": 977668, "author_name": "Marco Wu", "author_url": "", "post_date": "2020-08-19T16:26:09.970000", "content": "Thank you very much!! I already send an email to Phil Culliton. Hope they will give me a reply in 8 hours.
", "votes": 1, "replies": [] }, { "id": 977683, "author_name": "Marco Wu", "author_url": "", "post_date": "2020-08-19T16:36:36.297000", "content": "@stanleyjzheng My ground truth's csv_file was changed with my yolov5 distribution, so I named it as yolov5trainF5_onlyfix.csv. Perhaps it is one reason cause me error,but I think this should not be illegal.
\nThank you very much!!
", "votes": 1, "replies": [] }, { "id": 977728, "author_name": "Stanley Zheng", "author_url": "", "post_date": "2020-08-19T17:10:47.370000", "content": "I agree it definitely shouldn't, but it's possible they're just scanning for yolov5 in every notebook and invalidating them that way (more likely it's way more advanced or done by hand though, we'll see). Best of luck with your email.
", "votes": 1, "replies": [] }, { "id": 977754, "author_name": "Marco Wu", "author_url": "", "post_date": "2020-08-19T17:30:34.373000", "content": "Thanks!! Wish you can get great private score!
", "votes": 1, "replies": [] } ] }, { "id": 977628, "author_name": "Manoj Prabhakar", "author_url": "", "post_date": "2020-08-19T15:51:05.543000", "content": "@marcowu - I dont see any prohibited libraries as well. My only guess is it would have taken more than 6 hrs for the execution to complete.
", "votes": 1, "replies": [ { "id": 977639, "author_name": "Marco Wu", "author_url": "", "post_date": "2020-08-19T15:56:35.643000", "content": "My one submission was only running inference, not pseudo labeling.
", "votes": 1, "replies": [] } ] }, { "id": 977600, "author_name": "Marco Wu", "author_url": "", "post_date": "2020-08-19T15:34:20.540000", "content": "@philculliton Why this notebook even could get score,but I can not?
\n", "votes": 1, "replies": [] }, { "id": 977690, "author_name": "bendang", "author_url": "", "post_date": "2020-08-19T16:40:41.733000", "content": "Maybe because you are using RAdam in your dataset (not sure). I actually got more boost using RAdam but didn't use it as it was nowhere mentions in the external data thread. I even asked as a discussion topic but ended with zero response.
\nThis is just my speculation, I might be wrong here.
Thanks your reply!! My radam.py license is AP 2.0 in here. It is very disappoint that no kaggle team reply in your discussion. Because my one submission was only running inference, did not use radam. If it truly caused me out LB,I will hard to accept.
", "votes": 1, "replies": [] }, { "id": 977820, "author_name": "bendang", "author_url": "", "post_date": "2020-08-19T18:09:28.213000", "content": "I see that you did not use it in your inference code. If this is the reason being invalidated then I think the Kaggle team needs to reconsider your submission.
\nIt's best to reach out to the Kaggle team and clarify it. Hope they will restore your submission. Best of luck.