@abhishek has released a bunch of youtube videos for tackling Bengali.AI challenge. It's great to see how a grandmaster is approaching the competition and how he tackles the various issue.
\n\nDo check it out here:
\n\nPart1: https://youtu.be/8J5Q4mEzRtY\nPart2: https://youtu.be/uZalt-weQMM
", "messages": [ { "id": 753910, "postDate": "2020-02-22T20:04:42.170Z", "content": "@abhishek has released a bunch of youtube videos for tackling Bengali.AI challenge. It's great to see how a grandmaster is approaching the competition and how he tackles the various issue.
\n\nDo check it out here:
\n\nPart1: https://youtu.be/8J5Q4mEzRtY\nPart2: https://youtu.be/uZalt-weQMM
", "rawMarkdown": "@abhishek has released a bunch of youtube videos for tackling Bengali.AI challenge. It's great to see how a grandmaster is approaching the competition and how he tackles the various issue.\n\nDo check it out here:\n\nPart1: https://youtu.be/8J5Q4mEzRtY\nPart2: https://youtu.be/uZalt-weQMM", "votes": 6 }, { "id": 754066, "postDate": "2020-02-23T02:27:22.030Z", "content": "Great tutorial! @abhishek I think it worths to watch for more than one time.
", "rawMarkdown": "Great tutorial! @abhishek I think it worths to watch for more than one time.", "votes": 1 }, { "id": 753929, "postDate": "2020-02-22T20:41:38.513Z", "content": "Hi @kurianbenoy, hi @abhishek,
\n\nI saw this popping up in me Youtube-feat as well. \nI put a link in the \"External Data disclosure\" thread here? Since there is a lot of competition related pytorch kungfu going on in the videos.\nI really like the potcast. But IMHO publicly available competition related work should be mentioned either in the \"external Data\" thread or in a kaggle kernel during an ongoing competition. So people can find it easly.
", "rawMarkdown": "Hi @kurianbenoy, hi @abhishek,\n\nI saw this popping up in me Youtube-feat as well. \nI put a link in the \"External Data disclosure\" thread [here](https://www.kaggle.com/c/bengaliai-cv19/discussion/122396#753925)? Since there is a lot of competition related pytorch kungfu going on in the videos.\nI really like the potcast. But IMHO publicly available competition related work should be mentioned either in the \"external Data\" thread or in a kaggle kernel during an ongoing competition. So people can find it easly.", "votes": 1, "replies": [ { "id": 753937, "postDate": "2020-02-22T20:51:27.653Z", "content": "Hi @joatom \nexternal data disclosure thread is only for the external data that you use in the model, iirc :)
\n\nit seems the links have been posted a couple of times already today om the discussion and since its on YouTube, there is no private sharing. everything is public :)
", "rawMarkdown": "Hi @joatom \nexternal data disclosure thread is only for the external data that you use in the model, iirc :)\n\nit seems the links have been posted a couple of times already today om the discussion and since its on YouTube, there is no private sharing. everything is public :)", "votes": 1 }, { "id": 753953, "postDate": "2020-02-22T21:18:45.117Z", "content": "Hi @abhishek,\nI deleted my post in the external data disclosure thread. Since my framework is roughly build anyways and as far as I could see by quickly stepping through your videos there is no secrete magic that is not been mentioned in other kernels besides a well structured framework.\nMaybe you'd consider making a discussion topic yourself when you do a video on an ongoing competition next time. We had discussions in other competitions when people where sharing in forums or user groups.\nAnyway, no big deal ;) and the videos are certainly valuable and on my list for my commuting lecture next week.
", "rawMarkdown": "Hi @abhishek,\nI deleted my post in the external data disclosure thread. Since my framework is roughly build anyways and as far as I could see by quickly stepping through your videos there is no secrete magic that is not been mentioned in other kernels besides a well structured framework.\nMaybe you'd consider making a discussion topic yourself when you do a video on an ongoing competition next time. We had discussions in other competitions when people where sharing in forums or user groups.\nAnyway, no big deal ;) and the videos are certainly valuable and on my list for my commuting lecture next week.", "votes": 1 } ] }, { "id": 754970, "postDate": "2020-02-24T09:36:01.633Z", "content": "Oh, I'll watch that tonight
", "rawMarkdown": "Oh, I'll watch that tonight" }, { "id": 754952, "postDate": "2020-02-24T09:07:09.683Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 754066, "author_name": "DeepInvolution", "author_url": "", "post_date": "2020-02-23T02:27:22.030000", "content": "Great tutorial! @abhishek I think it worths to watch for more than one time.
", "votes": 1, "replies": [] }, { "id": 753929, "author_name": "Jo Tom", "author_url": "", "post_date": "2020-02-22T20:41:38.513000", "content": "Hi @kurianbenoy, hi @abhishek,
\n\nI saw this popping up in me Youtube-feat as well. \nI put a link in the \"External Data disclosure\" thread here? Since there is a lot of competition related pytorch kungfu going on in the videos.\nI really like the potcast. But IMHO publicly available competition related work should be mentioned either in the \"external Data\" thread or in a kaggle kernel during an ongoing competition. So people can find it easly.
", "votes": 1, "replies": [ { "id": 753937, "author_name": "Abhishek Thakur", "author_url": "", "post_date": "2020-02-22T20:51:27.653000", "content": "Hi @joatom \nexternal data disclosure thread is only for the external data that you use in the model, iirc :)
\n\nit seems the links have been posted a couple of times already today om the discussion and since its on YouTube, there is no private sharing. everything is public :)
", "votes": 1, "replies": [] }, { "id": 753953, "author_name": "Jo Tom", "author_url": "", "post_date": "2020-02-22T21:18:45.117000", "content": "Hi @abhishek,\nI deleted my post in the external data disclosure thread. Since my framework is roughly build anyways and as far as I could see by quickly stepping through your videos there is no secrete magic that is not been mentioned in other kernels besides a well structured framework.\nMaybe you'd consider making a discussion topic yourself when you do a video on an ongoing competition next time. We had discussions in other competitions when people where sharing in forums or user groups.\nAnyway, no big deal ;) and the videos are certainly valuable and on my list for my commuting lecture next week.
", "votes": 1, "replies": [] } ] }, { "id": 754970, "author_name": "Tian", "author_url": "", "post_date": "2020-02-24T09:36:01.633000", "content": "Oh, I'll watch that tonight
", "votes": 0, "replies": [] }, { "id": 754952, "author_name": "", "author_url": "", "post_date": "2020-02-24T09:07:09.683000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "753910": "@abhishek has released a bunch of youtube videos for tackling Bengali.AI challenge. It's great to see how a grandmaster is approaching the competition and how he tackles the various issue.\n\nDo check it out here:\n\nPart1: https://youtu.be/8J5Q4mEzRtY\nPart2: https://youtu.be/uZalt-weQMM", "754066": "Great tutorial! @abhishek I think it worths to watch for more than one time.", "753929": "Hi @kurianbenoy, hi @abhishek,\n\nI saw this popping up in me Youtube-feat as well. \nI put a link in the \"External Data disclosure\" thread [here](https://www.kaggle.com/c/bengaliai-cv19/discussion/122396#753925)? Since there is a lot of competition related pytorch kungfu going on in the videos.\nI really like the potcast. But IMHO publicly available competition related work should be mentioned either in the \"external Data\" thread or in a kaggle kernel during an ongoing competition. So people can find it easly.", "754970": "Oh, I'll watch that tonight", "754952": "" } }