I accidentally reproduced your labeling order. \nOf course I didn't achieve current score using neural networks.\nI hope it will not break this challenge and we can continue. But it means that everyone can use that info for validation at least.
", "messages": [ { "id": "773821", "postDate": "03/16/2020 00:08:55", "content": "I accidentally reproduced your labeling order. \nOf course I didn't achieve current score using neural networks.\nI hope it will not break this challenge and we can continue. But it means that everyone can use that info for validation at least.
", "rawMarkdown": "I accidentally reproduced your labeling order. \nOf course I didn't achieve current score using neural networks.\nI hope it will not break this challenge and we can continue. But it means that everyone can use that info for validation at least.", "votes": null }, { "id": "773823", "postDate": "03/16/2020 00:14:57", "content": "I have updated the test annotation. I hope you won't distribute the old one. I will contact Kaggle to remove your submission. Thanks!
", "rawMarkdown": "I have updated the test annotation. I hope you won't distribute the old one. I will contact Kaggle to remove your submission. Thanks!", "votes": null }, { "id": "774865", "postDate": "03/16/2020 02:30:56", "content": "We will collect new test data soon. Please don't use the old test data for training. Thank you!
", "rawMarkdown": "We will collect new test data soon. Please don't use the old test data for training. Thank you!", "votes": null }, { "id": "774879", "postDate": "03/16/2020 02:51:08", "content": "Does this mean we can not submit solutions until the new test data are collected?
", "rawMarkdown": "Does this mean we can not submit solutions until the new test data are collected?", "votes": null }, { "id": "774884", "postDate": "03/16/2020 02:56:26", "content": "We will pause the competition and get new test data soon. Meanwhile, you can still work on your algorithm and use validation data to see the performance. Sorry for my mistake.
", "rawMarkdown": "We will pause the competition and get new test data soon. Meanwhile, you can still work on your algorithm and use validation data to see the performance. Sorry for my mistake.", "votes": null }, { "id": "774895", "postDate": "03/16/2020 03:25:59", "content": "One more thing... I was going to download the data today, but it seems the links for train/val on the github pages are gone?
", "rawMarkdown": "One more thing... I was going to download the data today, but it seems the links for train/val on the [github pages](https://github.com/cvl-umass/semi-inat-2020) are gone?", "votes": null }, { "id": "774898", "postDate": "03/16/2020 03:32:23", "content": "You can download via Kaggle for now. I'm now pausing the competition.
", "rawMarkdown": "You can download via Kaggle for now. I'm now pausing the competition.", "votes": null }, { "id": "774899", "postDate": "03/16/2020 03:35:00", "content": "Thanks! But I still can't find annotations.tar.gz on kaggle...\nI mean, they seem to have a directory name indicating the class, do I have to select val images by myself?
", "rawMarkdown": "Thanks! But I still can't find annotations.tar.gz on kaggle...\nI mean, they seem to have a directory name indicating the class, do I have to select val images by myself?", "votes": null }, { "id": "774904", "postDate": "03/16/2020 03:40:59", "content": "The annotations have problems so I removed them for now. We will collect new data soon. Meanwhile you can use the last ten images (sorted by the filenames) in each folder as val.
", "rawMarkdown": "The annotations have problems so I removed them for now. We will collect new data soon. Meanwhile you can use the last ten images (sorted by the filenames) in each folder as val.", "votes": null }, { "id": "774905", "postDate": "03/16/2020 03:42:16", "content": "Ok, I understand, thank you.
", "rawMarkdown": "Ok, I understand, thank you.", "votes": null } ], "comments": [ { "id": 773823, "author_name": "jcfredsu", "author_url": "", "post_date": "03/16/2020 00:14:57", "content": "I have updated the test annotation. I hope you won't distribute the old one. I will contact Kaggle to remove your submission. Thanks!
", "votes": null, "replies": [] }, { "id": 774865, "author_name": "jcfredsu", "author_url": "", "post_date": "03/16/2020 02:30:56", "content": "We will collect new test data soon. Please don't use the old test data for training. Thank you!
", "votes": null, "replies": [] }, { "id": 774879, "author_name": "voldemortx", "author_url": "", "post_date": "03/16/2020 02:51:08", "content": "Does this mean we can not submit solutions until the new test data are collected?
", "votes": null, "replies": [ { "id": 774884, "author_name": "jcfredsu", "author_url": "", "post_date": "03/16/2020 02:56:26", "content": "We will pause the competition and get new test data soon. Meanwhile, you can still work on your algorithm and use validation data to see the performance. Sorry for my mistake.
", "votes": null, "replies": [] }, { "id": 774895, "author_name": "voldemortx", "author_url": "", "post_date": "03/16/2020 03:25:59", "content": "One more thing... I was going to download the data today, but it seems the links for train/val on the github pages are gone?
", "votes": null, "replies": [] }, { "id": 774898, "author_name": "jcfredsu", "author_url": "", "post_date": "03/16/2020 03:32:23", "content": "You can download via Kaggle for now. I'm now pausing the competition.
", "votes": null, "replies": [] }, { "id": 774899, "author_name": "voldemortx", "author_url": "", "post_date": "03/16/2020 03:35:00", "content": "Thanks! But I still can't find annotations.tar.gz on kaggle...\nI mean, they seem to have a directory name indicating the class, do I have to select val images by myself?
", "votes": null, "replies": [] }, { "id": 774904, "author_name": "jcfredsu", "author_url": "", "post_date": "03/16/2020 03:40:59", "content": "The annotations have problems so I removed them for now. We will collect new data soon. Meanwhile you can use the last ten images (sorted by the filenames) in each folder as val.
", "votes": null, "replies": [] }, { "id": 774905, "author_name": "voldemortx", "author_url": "", "post_date": "03/16/2020 03:42:16", "content": "Ok, I understand, thank you.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "773821": "I accidentally reproduced your labeling order. \nOf course I didn't achieve current score using neural networks.\nI hope it will not break this challenge and we can continue. But it means that everyone can use that info for validation at least.", "773823": "I have updated the test annotation. I hope you won't distribute the old one. I will contact Kaggle to remove your submission. Thanks!", "774865": "We will collect new test data soon. Please don't use the old test data for training. Thank you!", "774879": "Does this mean we can not submit solutions until the new test data are collected?", "774884": "We will pause the competition and get new test data soon. Meanwhile, you can still work on your algorithm and use validation data to see the performance. Sorry for my mistake.", "774895": "One more thing... I was going to download the data today, but it seems the links for train/val on the [github pages](https://github.com/cvl-umass/semi-inat-2020) are gone?", "774898": "You can download via Kaggle for now. I'm now pausing the competition.", "774899": "Thanks! But I still can't find annotations.tar.gz on kaggle...\nI mean, they seem to have a directory name indicating the class, do I have to select val images by myself?", "774904": "The annotations have problems so I removed them for now. We will collect new data soon. Meanwhile you can use the last ten images (sorted by the filenames) in each folder as val.", "774905": "Ok, I understand, thank you." }, "source": "meta" }