Hi Everyone, \nI have been doing a podcast for the past month: Chai Time Data Science, available both as video, audio (Blog post releases TBD) with 2 weekly episodes being published every week.
\n\nThis is V2.0 (if I may): I've been lucky enough to have interviewed many great Kagglers earlier as well. However, I thought its slightly unwise to call it a community-driven podcast since I wasn't exactly having direct involvement of the community. Earlier I requested @shaz13 for a similar AMA setup and that went really well (I'm still editing the episode. Turns out I'm slow with pipelines that are GPU Intensive even outside of the DL World :D)
\n\nWith the same approach in mind, I requested @antorsae to also allow me to collect questions and do an AMA and he has kindly agreed.
\n\nThe recording will happen this Sunday, 1st Sept evening (IST), I will take note of all the questions submitted here till then. Please feel free to post any/all questions here.
\n\nThe questions that I've drafted for him will be around the theme:
\n\n• His Kaggle Journey\n• Discussion of his current work outside Kaggle\n• Their 2nd place solution
\n\nPlease feel free to add questions related to the theme or to any topic. I'll try my best to include these in the interview which will be released as a Podcast later.
\n\nThank You!
", "messages": [ { "id": "613575", "postDate": "08/30/2019 16:21:39", "content": "Hi Everyone, \nI have been doing a podcast for the past month: Chai Time Data Science, available both as video, audio (Blog post releases TBD) with 2 weekly episodes being published every week.
\n\nThis is V2.0 (if I may): I've been lucky enough to have interviewed many great Kagglers earlier as well. However, I thought its slightly unwise to call it a community-driven podcast since I wasn't exactly having direct involvement of the community. Earlier I requested @shaz13 for a similar AMA setup and that went really well (I'm still editing the episode. Turns out I'm slow with pipelines that are GPU Intensive even outside of the DL World :D)
\n\nWith the same approach in mind, I requested @antorsae to also allow me to collect questions and do an AMA and he has kindly agreed.
\n\nThe recording will happen this Sunday, 1st Sept evening (IST), I will take note of all the questions submitted here till then. Please feel free to post any/all questions here.
\n\nThe questions that I've drafted for him will be around the theme:
\n\n• His Kaggle Journey\n• Discussion of his current work outside Kaggle\n• Their 2nd place solution
\n\nPlease feel free to add questions related to the theme or to any topic. I'll try my best to include these in the interview which will be released as a Podcast later.
\n\nThank You!
", "rawMarkdown": "Hi Everyone, \nI have been doing a podcast for the past month: [Chai Time Data Science](http://chaitimedatascience.com), available both as [video](https://www.youtube.com/playlist?list=PLLvvXm0q8zUbiNdoIazGzlENMXvZ9bd3x), [audio](https://anchor.fm/chaitimedatascience) (Blog post releases TBD) with 2 weekly episodes being published every week. \n\nThis is V2.0 (if I may): I've been [lucky enough to have interviewed many great Kagglers earlier as well](https://medium.com/dsnet/chai-time-data-science-show-announcement-bfaaf38df219). However, I thought its slightly unwise to call it a community-driven podcast since I wasn't exactly having direct involvement of the community. Earlier I requested @shaz13 for a similar AMA setup and that went really well (I'm still editing the episode. Turns out I'm slow with pipelines that are GPU Intensive even outside of the DL World :D) \n\nWith the same approach in mind, I requested @antorsae to also allow me to collect questions and do an AMA and he has kindly agreed. \n\nThe recording will happen this Sunday, 1st Sept evening (IST), I will take note of all the questions submitted here till then. Please feel free to post any/all questions here. \n\nThe questions that I've drafted for him will be around the theme:\n\n• His Kaggle Journey\n• Discussion of his current work outside Kaggle\n• Their 2nd place solution\n\nPlease feel free to add questions related to the theme or to any topic. I'll try my best to include these in the interview which will be released as a Podcast later.\n\nThank You!", "votes": null }, { "id": "613600", "postDate": "08/30/2019 16:44:46", "content": "Thanks for doing that @init27 ! Great initiative.
", "rawMarkdown": "Thanks for doing that @init27 ! Great initiative.", "votes": null }, { "id": "613619", "postDate": "08/30/2019 17:05:18", "content": "Why does he choose fast.ai instead of using native pytorch?
", "rawMarkdown": "Why does he choose fast.ai instead of using native pytorch?", "votes": null }, { "id": "613631", "postDate": "08/30/2019 17:16:42", "content": "This is interesting!
\n\nI have a couple of questions:\n- How do you choose in which competition to participate?\n- How do you balance time spend on competitions with life and work? :)
", "rawMarkdown": "This is interesting!\n\nI have a couple of questions:\n- How do you choose in which competition to participate?\n- How do you balance time spend on competitions with life and work? :)", "votes": null }, { "id": "613644", "postDate": "08/30/2019 17:29:07", "content": "1.Great how can Some one develop the Good Habit of Progressing on it each competition.\n2.How can Africans guys also have a Breakthrough dealing with kaggle Grandmaster.\n3.how to approach Different ML and DL problems
", "rawMarkdown": "1.Great how can Some one develop the Good Habit of Progressing on it each competition.\n2.How can Africans guys also have a Breakthrough dealing with kaggle Grandmaster.\n3.how to approach Different ML and DL problems", "votes": null }, { "id": "613807", "postDate": "08/30/2019 20:39:41", "content": "Some questions I have in mind as well:
\n- Can you describe your high-level approach when competing in a Kaggle competition? What is your Machine Learning workflow?
\n- What advice would you have for those currently learning Machine Learning and Data Science?
\n- What computing resources would you recommend to train gold-medal models in a reasonable amount of time? What is the minimal amount of computing resources to be able to iterate and learn fast for a given problem?
Can you please explain in detail the feature engineering that you used?\nWhat are the things that helped you increase your score gradually and which strategy was most effective?
", "rawMarkdown": "Can you please explain in detail the feature engineering that you used?\nWhat are the things that helped you increase your score gradually and which strategy was most effective?", "votes": null }, { "id": "614401", "postDate": "08/31/2019 12:49:04", "content": "How do track your experiments?
", "rawMarkdown": "How do track your experiments?", "votes": null }, { "id": "614650", "postDate": "08/31/2019 20:53:24", "content": "This looks great Sanyam; I'm looking forward to the podcast. Here are a few questions: How did you get started in deep learning? What can be done to get deep learning included in data science boot camps? How has what you have learnt taking part in Kaggle competitions impacted your professional life?
", "rawMarkdown": "This looks great Sanyam; I'm looking forward to the podcast. Here are a few questions: How did you get started in deep learning? What can be done to get deep learning included in data science boot camps? How has what you have learnt taking part in Kaggle competitions impacted your professional life?", "votes": null }, { "id": "614967", "postDate": "09/01/2019 09:53:10", "content": "I have also some questions:
\n\nThanks for doing this! 👍
", "rawMarkdown": "I have also some questions:\n\n1. What would be the first idea in his mind about feature selection?\n2. (Could be linked to the first question,) how to best allocate the resources (e.g. time), where to focus, if only limited resources are available?\n3. What would be his opinions / comments on AutoML?\n4. What would be the next big thing in data science?\n\nThanks for doing this! 👍", "votes": null }, { "id": "615166", "postDate": "09/01/2019 14:55:52", "content": "Thanks a lot, everyone for posting the questions here! \nThe call with Andres is about to start in 4 minutes and I've tried to include all of the questions posted until now. I won't be able to take the questions posted afterwards but I wanted to thank everyone for asking amazing questions and posting them here on short notice. \nI'll post a comment once the episode is out! (Meanwhile, do checkout the other episodes and let me know how do you like them)
", "rawMarkdown": "Thanks a lot, everyone for posting the questions here! \nThe call with Andres is about to start in 4 minutes and I've tried to include all of the questions posted until now. I won't be able to take the questions posted afterwards but I wanted to thank everyone for asking amazing questions and posting them here on short notice. \nI'll post a comment once the episode is out! (Meanwhile, do checkout the other episodes and let me know how do you like them)", "votes": null }, { "id": "616183", "postDate": "09/02/2019 20:03:42", "content": "Thanks Sanyam it was great to do the interview which I enjoyed a lot!
", "rawMarkdown": "Thanks Sanyam it was great to do the interview which I enjoyed a lot!", "votes": null }, { "id": "662868", "postDate": "11/01/2019 05:47:36", "content": "Thank you so much for doing the interview @antorsae! It was really an honor to talk to you, I enjoyed re-watching the interview a lot too
\n\nThe interview is live: \n- Video: https://www.youtube.com/watch?v=sqo2h7aYPPk&list=PLLvvXm0q8zUbiNdoIazGzlENMXvZ9bd3x&index=2\n- Audio: https://anchor.fm/dashboard/episode/e8h0tj
\n\nThanks again to Andres and to everyone for submitting their questions! 😄
", "rawMarkdown": "Thank you so much for doing the interview @antorsae! It was really an honor to talk to you, I enjoyed re-watching the interview a lot too\n\nThe interview is live: \n- Video: https://www.youtube.com/watch?v=sqo2h7aYPPk&list=PLLvvXm0q8zUbiNdoIazGzlENMXvZ9bd3x&index=2\n- Audio: https://anchor.fm/dashboard/episode/e8h0tj\n\nThanks again to Andres and to everyone for submitting their questions! 😄", "votes": null } ], "comments": [ { "id": 613600, "author_name": "michelml", "author_url": "", "post_date": "08/30/2019 16:44:46", "content": "Thanks for doing that @init27 ! Great initiative.
", "votes": null, "replies": [] }, { "id": 613619, "author_name": "returnofsputnik", "author_url": "", "post_date": "08/30/2019 17:05:18", "content": "Why does he choose fast.ai instead of using native pytorch?
", "votes": null, "replies": [] }, { "id": 613631, "author_name": "artgor", "author_url": "", "post_date": "08/30/2019 17:16:42", "content": "This is interesting!
\n\nI have a couple of questions:\n- How do you choose in which competition to participate?\n- How do you balance time spend on competitions with life and work? :)
", "votes": null, "replies": [ { "id": 613807, "author_name": "michelml", "author_url": "", "post_date": "08/30/2019 20:39:41", "content": "Some questions I have in mind as well:
\n- Can you describe your high-level approach when competing in a Kaggle competition? What is your Machine Learning workflow?
\n- What advice would you have for those currently learning Machine Learning and Data Science?
\n- What computing resources would you recommend to train gold-medal models in a reasonable amount of time? What is the minimal amount of computing resources to be able to iterate and learn fast for a given problem?
1.Great how can Some one develop the Good Habit of Progressing on it each competition.\n2.How can Africans guys also have a Breakthrough dealing with kaggle Grandmaster.\n3.how to approach Different ML and DL problems
", "votes": null, "replies": [] }, { "id": 614326, "author_name": "vikrant06", "author_url": "", "post_date": "08/31/2019 11:04:11", "content": "Can you please explain in detail the feature engineering that you used?\nWhat are the things that helped you increase your score gradually and which strategy was most effective?
", "votes": null, "replies": [] }, { "id": 614401, "author_name": "arnabbiswas1", "author_url": "", "post_date": "08/31/2019 12:49:04", "content": "How do track your experiments?
", "votes": null, "replies": [] }, { "id": 614650, "author_name": "alisondavey", "author_url": "", "post_date": "08/31/2019 20:53:24", "content": "This looks great Sanyam; I'm looking forward to the podcast. Here are a few questions: How did you get started in deep learning? What can be done to get deep learning included in data science boot camps? How has what you have learnt taking part in Kaggle competitions impacted your professional life?
", "votes": null, "replies": [] }, { "id": 614967, "author_name": "giovawang", "author_url": "", "post_date": "09/01/2019 09:53:10", "content": "I have also some questions:
\n\nThanks for doing this! 👍
", "votes": null, "replies": [] }, { "id": 615166, "author_name": "init27", "author_url": "", "post_date": "09/01/2019 14:55:52", "content": "Thanks a lot, everyone for posting the questions here! \nThe call with Andres is about to start in 4 minutes and I've tried to include all of the questions posted until now. I won't be able to take the questions posted afterwards but I wanted to thank everyone for asking amazing questions and posting them here on short notice. \nI'll post a comment once the episode is out! (Meanwhile, do checkout the other episodes and let me know how do you like them)
", "votes": null, "replies": [] }, { "id": 616183, "author_name": "antorsae", "author_url": "", "post_date": "09/02/2019 20:03:42", "content": "Thanks Sanyam it was great to do the interview which I enjoyed a lot!
", "votes": null, "replies": [ { "id": 662868, "author_name": "init27", "author_url": "", "post_date": "11/01/2019 05:47:36", "content": "Thank you so much for doing the interview @antorsae! It was really an honor to talk to you, I enjoyed re-watching the interview a lot too
\n\nThe interview is live: \n- Video: https://www.youtube.com/watch?v=sqo2h7aYPPk&list=PLLvvXm0q8zUbiNdoIazGzlENMXvZ9bd3x&index=2\n- Audio: https://anchor.fm/dashboard/episode/e8h0tj
\n\nThanks again to Andres and to everyone for submitting their questions! 😄
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "613575": "Hi Everyone, \nI have been doing a podcast for the past month: [Chai Time Data Science](http://chaitimedatascience.com), available both as [video](https://www.youtube.com/playlist?list=PLLvvXm0q8zUbiNdoIazGzlENMXvZ9bd3x), [audio](https://anchor.fm/chaitimedatascience) (Blog post releases TBD) with 2 weekly episodes being published every week. \n\nThis is V2.0 (if I may): I've been [lucky enough to have interviewed many great Kagglers earlier as well](https://medium.com/dsnet/chai-time-data-science-show-announcement-bfaaf38df219). However, I thought its slightly unwise to call it a community-driven podcast since I wasn't exactly having direct involvement of the community. Earlier I requested @shaz13 for a similar AMA setup and that went really well (I'm still editing the episode. Turns out I'm slow with pipelines that are GPU Intensive even outside of the DL World :D) \n\nWith the same approach in mind, I requested @antorsae to also allow me to collect questions and do an AMA and he has kindly agreed. \n\nThe recording will happen this Sunday, 1st Sept evening (IST), I will take note of all the questions submitted here till then. Please feel free to post any/all questions here. \n\nThe questions that I've drafted for him will be around the theme:\n\n• His Kaggle Journey\n• Discussion of his current work outside Kaggle\n• Their 2nd place solution\n\nPlease feel free to add questions related to the theme or to any topic. I'll try my best to include these in the interview which will be released as a Podcast later.\n\nThank You!", "613600": "Thanks for doing that @init27 ! Great initiative.", "613619": "Why does he choose fast.ai instead of using native pytorch?", "613631": "This is interesting!\n\nI have a couple of questions:\n- How do you choose in which competition to participate?\n- How do you balance time spend on competitions with life and work? :)", "613644": "1.Great how can Some one develop the Good Habit of Progressing on it each competition.\n2.How can Africans guys also have a Breakthrough dealing with kaggle Grandmaster.\n3.how to approach Different ML and DL problems", "613807": "Some questions I have in mind as well: \n- Can you describe your high-level approach when competing in a Kaggle competition? What is your Machine Learning workflow? \n- What advice would you have for those currently learning Machine Learning and Data Science? \n- What computing resources would you recommend to train gold-medal models in a reasonable amount of time? What is the minimal amount of computing resources to be able to iterate and learn fast for a given problem?", "614326": "Can you please explain in detail the feature engineering that you used?\nWhat are the things that helped you increase your score gradually and which strategy was most effective?", "614401": "How do track your experiments?", "614650": "This looks great Sanyam; I'm looking forward to the podcast. Here are a few questions: How did you get started in deep learning? What can be done to get deep learning included in data science boot camps? How has what you have learnt taking part in Kaggle competitions impacted your professional life?", "614967": "I have also some questions:\n\n1. What would be the first idea in his mind about feature selection?\n2. (Could be linked to the first question,) how to best allocate the resources (e.g. time), where to focus, if only limited resources are available?\n3. What would be his opinions / comments on AutoML?\n4. What would be the next big thing in data science?\n\nThanks for doing this! 👍", "615166": "Thanks a lot, everyone for posting the questions here! \nThe call with Andres is about to start in 4 minutes and I've tried to include all of the questions posted until now. I won't be able to take the questions posted afterwards but I wanted to thank everyone for asking amazing questions and posting them here on short notice. \nI'll post a comment once the episode is out! (Meanwhile, do checkout the other episodes and let me know how do you like them)", "616183": "Thanks Sanyam it was great to do the interview which I enjoyed a lot!", "662868": "Thank you so much for doing the interview @antorsae! It was really an honor to talk to you, I enjoyed re-watching the interview a lot too\n\nThe interview is live: \n- Video: https://www.youtube.com/watch?v=sqo2h7aYPPk&list=PLLvvXm0q8zUbiNdoIazGzlENMXvZ9bd3x&index=2\n- Audio: https://anchor.fm/dashboard/episode/e8h0tj\n\nThanks again to Andres and to everyone for submitting their questions! 😄" }, "source": "meta" }