Jeremy Howard Interviews Kaggle Grandmaster Sanyam Bhutani

‪@ChaiTimeDataScience‬ (CTDS) becomes "Coffee Time Data Science" for a one-time special episode! The genesis for this interview was this twitter thread: / 1503920390447730690 . For many more CTDS interviews, check out the feeds at chaitimedatascience.com/ .
Links from the show:
CTDS Show Channel: / chaitimedatascience
Audio Version: anchor.fm/chaitimedatascience/
fastai Course: course.fast.ai
How not to do fastai: / how-not-to-do-fast-ai-...
Meta Learning Book: radekosmulski.gumroad.com/l/l...
Rachel’s advice on Blogging: / why-you-yes-you-should...
In the frame:
Jeremy Howard: / jeremyphoward
Sanyam Bhutani: / bhutanisanyam1
OUTLINE:
0:00 - Intro
2:40 - Reaching Kaggle Grandmaster Tier
4:46 - Current Work
6:30 - University Education
12:00 - Signing up for fastai
15:49 - Failures with fastai
18:29 - Tenacity
23:40 - Content Creation
33:10 - Sharing your work
36:09 - Educational Content
41:49 - Failing Google AI Residency
46:34 - Starting the podcast
48:19 - Question to Jeremy
52:40 - Thank you

[@mayankpj]

❤ Only Jeremy can do this! I'm so grateful for having you Jeremy as a teacher who I've never met in person but you're one of the best teachers I've learned a lot from .... like many others :)

[@matteo4318]

Such a great interview! 🔥 I had the great opportunity to both learn from and collaborate with Sanyam, and I can 100% confirm he's an amazing person. Keep up the good work Sanyam and thanks Jeremy for organizing the interview.

[@mahe8116]

I still remember when I discovered FastAI ..how it not only changed the way I learned ML...it even completely changed the way I approached software engineering! Thank you very much @Jeremy

[@wilcovanginkel9256]

Thanks for sharing this podcast. I really enjoyed listening to it and found the conversation refreshing in terms of the questions and answers. I am a big fan of both Kaggle and in particular fast.ai. It is great to see the younger generation stepping up to the plate, getting inspired, and trying to make the world a better place. Well done & keep up the good work!

[@user-qj4zr1pj9y]

Such Humble person Jeremy is !!! Congrats Sanyam on being interviewed by your role model !!

[@Yenrabbit]

Great of you to do this Jeremy, what a lovely idea! Nice to a new Grandmaster experiencing the other side of the microphone for a change :)

[@MachineLearningStreetTalk]

Amazing! Really appreciated the frank and honest exchange, great conversation

[@ScottSenkeresty]

"Enjoy the process of SHOWING UP. Every day". Legend.

[@carlolepelaars]

A treasure trove of wisdom. Thank you both for sharing these insights!

[@f_ipaforschung5514]

I have followed both of you for the last 5 years. Great to see both of you come together :). Learned a lot from this episode.

[@coreygallon]

This was brilliant, fellows! Sanyam is so very humble. Thank you, both, for sharing!

[@pursh2002]

enjoyed every bit of it ... Thank you

[@TheLokiGT]

My compliments, Sanyam! ;)

[@meghrajpardesi4181]

Hi @Jeremy Howard thanks for video , also the way you started the video had a smile on my face thanks for that too.

[@adityabhatt04]

Wow! Really loved this one.

[@chiragsharma9430]

It's nice to see Sanyam being interviewed this time. Sanyam is looking very happy to be interviewed by Jeremy. 😃😃

[@ibrahimn628]

I came across fast Ai a few months ago and after reading the description of the project and Jeremy’s philosophy of teaching/studying, it’s called top-down as opposed to bottom up, I decided to take the fats Ai courses. However, I never got around staring it because I was busy with school and work and I didn’t want to start it and then stopped it after a few weeks, which is a bad habit that I’m struggling to get rid off. I’ll start the fast Ai courses 2 weeks from now, shooting for 1 hour every weekday and 3 hours for weekends. My goals is to be persistent and not to skip the course 2 days in a raw. Thank you Jeremy for the good work you are doing!

[@colew3841]

This is amazing! Keep it up Jeremy 👍

[@dylanloader3869]

Such a great discussion with 2 of my favourite content creators in the space. - I thought Sanyam's discussion of the tutorial grinding without 'creating' was really well put as I think it's a common problem in ML/DL, especially with how common it is to get stuck in the cycle of getting certificates when job hunting. - As much as it's a bummer Sanyam didn't get into the Google AI Residency, I'm sort of glad because his personality and method of instruction is so unique that it would have been a shame if he wasn't as public facing as he has been at H2O and WANDB. Great content as always!

[@ashish-blessings]

Thank you 😊

[@giancarlovilchezquispe1752]

Sanyam emotion and humility represents me. I admire you both, good interview.

[@talk2yuvraj]

In case you came looking for nuggets of wisdom and experience from a fulfilling journey of both Jeremy and Saynam, feel free to what I have summarized from this interview for myself (cheers to both for doing this): - Good way of imparting education (A Book or A teacher in an institute) is what makes an experience worthwhile, not a specific college degree in a specific subject. - There is a huge significance of top-down approach of learning. We default to bottom-up style every time we get stuck and end up feeling like we need to get our basics good first. Just start working on something and learn that way. - Believing in the journey (a course, for example) and not thinking much about the end-goal can help you in staying tenacious or determined. Looking at the goal / a person as an end-goal, can overwhelm you. Just show-up every day and give yourself some months and it would be just fine! - Believe in the system - that working hard gives you good results. If you don't believe that this is true, then its going to be very rough. - People who keep coming back to some challenge, make it. - Create/write something that you yourself would wish to be out there. Try and fill gaps that you think do exist in the field you are learning (Ergo, this post :) ) - Give a try to something and you will fail sometimes, never try and you will fail a 100 percent of times. Like Jeremy felt it, you do somethings out of your comfort zone, you feel terrified while doing them and even exhausted after it. But you got to do it anyways, otherwise you will always fail if you won't give yourself chance to succeed. - Surround yourself with people who appreciate you, who value you for who you are and what you do. - People that you admire and look up to are not on some another level. They are normal people too, they make mistakes, they have setbacks too, but they redouble their efforts after that. You can also do "interesting" things that they are doing. It's not so difficult.

[@manamsetty2664]

Dude 🤯 thanks i usually go to the comments to feel the video thank you

[@8shailesh]

Really loved this interaction.

[@mehuljain2950]

I want to see more episodes like this 👍👍

[@Sagaofsr]

Love this

[@MayukhBose]

OMG hard relate with so much stuff you said about colleges in India!!

[@ashu-]

wow, coffee time!

[@prabhavkaula9697]

from chai time data science to coffee time data science :)

[@robosergTV]

Any ETA on the 2021 course?

[@arjunp2014]

hilarious intro

[@ZZmogus]

Where Kaggle Kernels are located on Kaggle website?

[@rajbanwait325]

HOMEWORK...