Daniel Bourke
Daniel Bourke
  • 307
  • 9 395 917
A Day in the Life of a Machine Learning Engineer (at a *small* startup)
A day in the life of a machine learning engineer at a small startup from Brisbane, Australia called Nutrify.
Nutrify uses computer vision to make learning about and tracking whole foods fun.
Links from video:
Download Nutrify on the App Store - apple.co/4ahM7Wc
Nutrify website - www.nutrify.app
What we learned from year of building LLMs part II - www.oreilly.com/radar/what-we-learned-from-a-year-of-building-with-llms-part-ii/
Apple Overton System for Machine Learning - machinelearning.apple.com/research/overton
Copying Tesla’s Data Engine for Food Images - ua-cam.com/video/02b-1spzLzc/v-deo.htmlsi=p7HBZ3dLaq6NGC57
My ML recommended resources - www.mrdbourke.com/ml-resources/
2020 Deep Learning PC Part List - au.pcpartpicker.com/list/8KXXjp
2024 Deep Learning PC Part List - au.pcpartpicker.com/list/xBvRdH
Other links:
Download Nutrify (my startup) - nutrify.app
Learn AI/ML (beginner-friendly course) - dbourke.link/ZTMMLcourse
Learn TensorFlow - dbourke.link/ZTMTFcourse
Learn PyTorch - dbourke.link/ZTMPyTorch
Read my novel Charlie Walks - www.charliewalks.com
Connect elsewhere:
Web - dbourke.link/web
X/Twitter - mrdbourke
Twitch - www.twitch.tv/mrdbourke
ArXiv channel (past streams) - dbourke.link/archive-channel
Get email updates on my work - dbourke.link/newsletter
Timestamps:
0:00 - intro
1:08 - morning session/breakfast
1:41 - reading
2:15 - data labelling
2:40 - model training
4:19 - bug fixing
4:34 - lunchtime
6:09 - afternoon session
6:51 - what we've been working on (for the day)
8:36 - ideas for the future
9:28 - Nutrify compute cluster in a closet
11:00 - Nvidia GPU speed comparison
12:00 - Nutrify's data flywheel
13:14 - end of day
14:34 - bloopers
Переглядів: 187 587

Відео

4 MLOps Lessons Learned from Bootstrapping an AI Startup (talk)
Переглядів 6 тис.2 місяці тому
I gave a talk about MLOps lessoned learned from building my brother and I's startup Nutrify (a computer vision powered app to help people learn about food). The talk originally happened in the MLOps Learners Discord and was livestreamed to UA-cam. I'm reposting my section here. MLOps stands for "Machine Learning Operations" and includes all of the parts around a machine learning model. For exam...
Your NVIDIA GPU can power a free ChatGPT (Chat with RTX)
Переглядів 10 тис.3 місяці тому
Chat with RTX is a ChatGPT-style application that uses Retrieval Augmented Generation (RAG) to allow you to chat with your own data on your own local NVIDIA GPU/PC. Links mentioned: Download Chat with RTX - www.nvidia.com/en-au/ai-on-rtx/chat-with-rtx-generative-ai/ NVIDIA RAG blog post - developer.nvidia.com/blog/rag-101-demystifying-retrieval-augmented-generation-pipelines/ My RAG tutorial (c...
Local Retrieval Augmented Generation (RAG) from Scratch (step by step tutorial)
Переглядів 63 тис.3 місяці тому
In this video we'll build a Retrieval Augmented Generation (RAG) pipeline to run locally from scratch. There are frameworks to do this such as LangChain and LlamaIndex, however, building from scratch means that you'll know all the parts of the puzzle. Specifically, we'll build NutriChat, a RAG pipeline that allows someone to ask questions of a 1200 page Nutrition Textbook PDF. Code on GitHub - ...
Using AI to build a Pokédex for Food
Переглядів 8 тис.4 місяці тому
Nutrify allows you to take a photo of food and learn about it. Consider it a Pokédex for food or Shazam for food. Download Nutrify on the App Store - apple.co/4ahM7Wc Nutrify website - www.nutrify.app Nutrify Instagram - nutrifyfoodapp Questions? - support@nutrify.app Other links: Learn AI/ML (beginner-friendly course) - dbourke.link/ZTMMLcourse Learn TensorFlow - dbourke.link/ZTM...
Apple M3 Machine Learning Speed Test (M1 Pro vs M3, M3 Pro, M3 Max)
Переглядів 165 тис.6 місяців тому
I put the latest Apple Silicon Macs (M3, M3 Pro, M3 Max) M3 series Macs through a series of machine learning speed tests with PyTorch and TensorFlow. Code on GitHub - github.com/mrdbourke/mac-ml-speed-test Blog post write up - www.mrdbourke.com/apple-m3-machine-learning-test/ Learn ML (taught by me) - www.mrdbourke.com/ml-courses/ Links mentioned: MLX framework by Apple - github.com/ml-explore/...
5 free resources to help you get a machine learning job
Переглядів 54 тис.Рік тому
This video goes through five of my favourite resources for getting a machine learning job. Blog post version of this video (with all links) - www.mrdbourke.com/get-a-machine-learning-job Links mentioned (in order): 1. Learn in Public by Swyx - www.swyx.io/learn-in-public 2. You Don't Really Need Another MOOC by Eugene Yan - eugeneyan.com/writing/you-dont-need-another-mooc/ 3. Start the job befo...
PyTorch 2.0 is here! Quick tutorial + NVIDIA RTX 4080 giveaway
Переглядів 15 тис.Рік тому
The official version PyTorch 2.0 just launched! In this video, I break down some of the updates and how you can use them in your PyTorch code. Sign up for NVIDIA GTC March 2023 - nvda.ws/3GUZygQ Details for giveaway - www.mrdbourke.com/pytorch-2 NVIDIA GTC sessions I'm going to: Compile and train with 43% speedup using PyTorch 2.0 - dbourke.link/nvidia-gtc-2023-pytorch-2 Fireside Chat with Ilya...
the day ai went everywhere (GPT-4 & Google PaLM API release)
Переглядів 9 тис.Рік тому
GPT-4 came out today along with new generative AI APIs from Google! And I think it's the momentum step to AI being in everything. Soon every app will be an AI app. This video dAIry (dAIry = AI diary) was recorded live with no edits/takes. Just raw thoughts and ideas for future Dan to look back on and laugh. Links: • See a fantastic technical report read through from @aiexplained-official - ua-c...
Copying Tesla's data engine, but with food images | nutrify #1
Переглядів 16 тис.Рік тому
Nutrify is an upcoming iOS app my brother and I are building to take a photo of food and learn about it. This video is the first in a series of videos documenting how it's getting built with a focus on the machine learning component. Links: Sign up for more - forms.gle/8ye134pvgEKWJQHd6 Nutrify website - nutrify.app Nutrify GitHub - github.com/mrdbourke/nutrify Blog post with more details - www...
Machine Learning Engineer Writes a Novel (without using AI)
Переглядів 4,7 тис.Рік тому
Buy Charlie Walks: A Novel - www.charliewalks.com Other links: Learn ML (beginner-friendly courses I teach) - www.mrdbourke.com/ml-courses/ ML courses/books I recommend - www.mrdbourke.com/ml-resources/ Read my novel Charlie Walks - www.charliewalks.com Connect elsewhere: Web - dbourke.link/web Twitter - mrdbourke Twitch - www.twitch.tv/mrdbourke ArXiv channel (past streams) - dbour...
Learn PyTorch for deep learning in a day. Literally.
Переглядів 1,5 млнРік тому
Welcome to the most beginner-friendly place on the internet to learn PyTorch for deep learning. All code on GitHub - dbourke.link/pt-github Ask a question - dbourke.link/pt-github-discussions Read the course materials online - learnpytorch.io Sign up for the full course on Zero to Mastery (20 hours more video) - dbourke.link/ZTMPyTorch Below are the timestamps/outline of the video. The video yo...
Setup Mac for Machine Learning with PyTorch in 11 minutes (works for all M1, M2)
Переглядів 73 тис.2 роки тому
Setup Mac for Machine Learning with PyTorch in 11 minutes (works for all M1, M2)
The Unofficial PyTorch Optimization Loop Song
Переглядів 23 тис.2 роки тому
The Unofficial PyTorch Optimization Loop Song
How to learn data science with Coursera in 2022 (beginner-friendly)
Переглядів 133 тис.2 роки тому
How to learn data science with Coursera in 2022 (beginner-friendly)
The most important article on programming you’ll ever read
Переглядів 13 тис.2 роки тому
The most important article on programming you’ll ever read
Apple's M1 Pro and M1 Max are faster than Google Colab (machine learning speed test)
Переглядів 61 тис.2 роки тому
Apple's M1 Pro and M1 Max are faster than Google Colab (machine learning speed test)
Setup Mac for Machine Learning with TensorFlow in 13 minutes (works for all M1, M2)
Переглядів 111 тис.2 роки тому
Setup Mac for Machine Learning with TensorFlow in 13 minutes (works for all M1, M2)
How to code neural networks without math*
Переглядів 8 тис.2 роки тому
How to code neural networks without math*
6 Techniques That Help Me Study Machine Learning Five Days Per Week
Переглядів 83 тис.2 роки тому
6 Techniques That Help Me Study Machine Learning Five Days Per Week
AI makes you an artist + Tesla's self-driving car updates | Machine Learning Monthly August 2021
Переглядів 6 тис.2 роки тому
AI makes you an artist Tesla's self-driving car updates | Machine Learning Monthly August 2021
Reading the first 42 pages of my first novel (Charlie Walks) out loud
Переглядів 3,9 тис.2 роки тому
Reading the first 42 pages of my first novel (Charlie Walks) out loud
I didn't eat food for 3 days, here's what happened
Переглядів 16 тис.2 роки тому
I didn't eat food for 3 days, here's what happened
Demo your ML models + learn NLP from the best (Hugging Face 🤗) | Machine Learning Monthly July 2021
Переглядів 15 тис.2 роки тому
Demo your ML models learn NLP from the best (Hugging Face 🤗) | Machine Learning Monthly July 2021
This book will help you get a job in machine learning (ML Interviews Book) | ML Monthly June 2021
Переглядів 32 тис.2 роки тому
This book will help you get a job in machine learning (ML Interviews Book) | ML Monthly June 2021
These TensorFlow updates are huuuuuuuuge! [Google I/O 2021] | Machine Learning Monthly May 2021
Переглядів 7 тис.3 роки тому
These TensorFlow updates are huuuuuuuuge! [Google I/O 2021] | Machine Learning Monthly May 2021
How the World's Biggest Companies Design Machine Learning Applications | ML Monthly April 2021
Переглядів 13 тис.3 роки тому
How the World's Biggest Companies Design Machine Learning Applications | ML Monthly April 2021
ResNets are back baby!!! ...and they're gone | Machine Learning Monthly March 2021
Переглядів 7 тис.3 роки тому
ResNets are back baby!!! ...and they're gone | Machine Learning Monthly March 2021
Learn TensorFlow and Deep Learning fundamentals with Python (code-first introduction) Part 2/2
Переглядів 380 тис.3 роки тому
Learn TensorFlow and Deep Learning fundamentals with Python (code-first introduction) Part 2/2
Learn TensorFlow and Deep Learning fundamentals with Python (code-first introduction) Part 1/2
Переглядів 810 тис.3 роки тому
Learn TensorFlow and Deep Learning fundamentals with Python (code-first introduction) Part 1/2

КОМЕНТАРІ

  • @jakubzboina7246
    @jakubzboina7246 Годину тому

    Very nice :D I'm training my LLMs and watching pros in action. Good luck ;)

  • @YUUKI_VLOG.10
    @YUUKI_VLOG.10 3 години тому

    かっこいい😎

  • @Crytoma
    @Crytoma 4 години тому

    Wholesome

  • @paziipa
    @paziipa 4 години тому

    Why not multiple gpus per machine?

  • @scott_strang
    @scott_strang 6 годин тому

    Hot dog. Not hot dog.

  • @ShareefDaana
    @ShareefDaana 6 годин тому

    talk to me i have advice for you man - in private

  • @EmilyClaire-z5k
    @EmilyClaire-z5k 9 годин тому

    this is so fun to watch and motivating that i feel like working with you guys or build an awesome team like yours

  • @Sinanmmd
    @Sinanmmd 10 годин тому

    I like your startup, could you please tell us about your PC which you are using that for model training? What are the specs and all that, thank you <3

  • @muzeroj173
    @muzeroj173 10 годин тому

    Maybe setup a remote environment, so you can train, debug your model and code by your desk.👍

  • @mitaskeledzija6269
    @mitaskeledzija6269 12 годин тому

    Can I join? I need a job lmao

  • @tirdad_ai
    @tirdad_ai 14 годин тому

    thank you daniel . this course is the best pytorch course in the internet.thank you very much🙏🙏🙏

  • @userasd360
    @userasd360 16 годин тому

    Isn't the ubuntu version a bit old? Would you upgrade it to the latest version?

  • @WalkAloneLive
    @WalkAloneLive 17 годин тому

    I recognize pizza dad! 😀

  • @user-ky9ku3wt7u
    @user-ky9ku3wt7u 19 годин тому

    Make video on phishing site detection using machine learning ,please

  • @santiagoroldan3163
    @santiagoroldan3163 23 години тому

    Daniel, thanks for the video! You could make a video about what it's like to create a startup, how to get customers and everything related to a startup.

  • @gubatron
    @gubatron День тому

    wonder if it'd be cheaper to do the training in AWS Cloud, and certainly cut down time from 11-12 hours, to 1-2 hours (10x).

    • @mrdbourke
      @mrdbourke День тому

      It could be, I haven’t experimented with it myself. Maybe could rent out a large cluster for an hour or so for training runs. But I like the convenience of having a GPU station on hand.

  • @aditipandey1769
    @aditipandey1769 День тому

    This video is one of the most authentic, wholesome and amazing vlog. Really felt the good vibes right through the screen. Loved it.

    • @mrdbourke
      @mrdbourke День тому

      Thank you! Glad you enjoyed it :)

  • @alit2086
    @alit2086 День тому

    Hotdog! Not Hotdog!

    • @mrdbourke
      @mrdbourke День тому

      Nutrify identifies hot dogs too!

  • @ThePCShow
    @ThePCShow День тому

    15:59:08 shoutout to the lovely cuckoo

    • @mrdbourke
      @mrdbourke День тому

      I knew there would be a bird in there somewhere! Thank you for pointing it out :)

  • @zainkhan2350
    @zainkhan2350 День тому

    No, my app is a see food ... like a food, that you can see.

    • @mrdbourke
      @mrdbourke День тому

      It does Seafood too! :P

  • @marcopedraza1268
    @marcopedraza1268 День тому

    How did you learn all that?

    • @mrdbourke
      @mrdbourke День тому

      Years of study + practice. See my earlier UA-cam videos for when I got started. Or maybe see here too: dbourke.link/aimastersdegree (I created my own AI Masters Degree and studied online for a year)

  • @user-gn5th9vv4k
    @user-gn5th9vv4k День тому

    Bro, I don't think that this is a small startup!!

    • @mrdbourke
      @mrdbourke День тому

      A ‘tiny’ startup? :P

  • @nvasudeva
    @nvasudeva День тому

    Just curious why you don't outsource manual labelling process . wouldn't that save you a lot of your's and your team time ? Is it because u don't trust the quality of the outsourcing or is it the budget issue ?

    • @mrdbourke
      @mrdbourke День тому

      Great idea! But most of our labelling is done by models. I personally spend ~1 hour per week on labelling. And my friend Joey comes around and helps for 4-5 hours per week (part-time). You are right on quality/budget. We do almost everything in house to save money and ensure quality. If our product is driven by a model, quality data is like quality code.

  • @yeabsiragashaw2099
    @yeabsiragashaw2099 День тому

    Truly Amazing , Keep it up !

  • @emerleite
    @emerleite День тому

    Witch pre-trained model are you using?

    • @mrdbourke
      @mrdbourke День тому

      We are using a “Next-ViT” backbone and then fine-tuning it on our own dataset for FoodVision and a “FastViT” backbone for food/not_food identification. Both of these are available on Hugging Face.

  • @thisizspeed
    @thisizspeed День тому

    Where did you get the bookshelf from?

    • @mrdbourke
      @mrdbourke День тому

      I got it from a furniture store about 25 minutes from my house (Freedom Furniture Brisbane)

  • @LyanQ
    @LyanQ День тому

    This was really a day in the life video, really appreciate your works.

    • @mrdbourke
      @mrdbourke День тому

      Thank you! Glad you enjoyed it :)

  • @jasonhall6526
    @jasonhall6526 День тому

    No one cares what is written on your whiteboard......

  • @mauroquiroga225
    @mauroquiroga225 День тому

    What a great video! You really hooked me throughout the entire video. Now I want to start learning ML asap 😂

    • @mrdbourke
      @mrdbourke День тому

      Thank you! Glad you enjoyed it! If you wanna learn ML, I’m biased but I’ve got some great beginner-friendly courses: www.mrdbourke.com/ml-courses/

  • @Tyheir
    @Tyheir День тому

    You look so happy I love it. I wanna get into ML don’t know where to start tho

    • @mrdbourke
      @mrdbourke День тому

      ML is quite the big field! I’d say one of the best places to get started is to learn Python programming and then start building some models of your own. I’ve got a blog post here about how I got started: www.mrdbourke.com/startaiml/

  • @buckyzona
    @buckyzona День тому

    Cool!

  • @Demonx99999
    @Demonx99999 День тому

    I hope that one day I will have a hard-working and harmonious team. Good work !

  • @avinashmehta371
    @avinashmehta371 День тому

    Thanks a lot Daniel!

    • @mrdbourke
      @mrdbourke День тому

      You’re welcome Avinashmehta!

  • @muhammadawais721
    @muhammadawais721 День тому

    This is very informative and practical Day of life video I ever seen. Nice Video Man😍😍

    • @mrdbourke
      @mrdbourke День тому

      Thank you my friend! I really appreciate it

  • @kitgary
    @kitgary День тому

    prSoftware engineer will be replaced by AI soon! We will be unemployed!

  • @FazriGading
    @FazriGading День тому

    Thank you for sharing, Daniel! I wonder that I could be cool like you someday. Wish me luck on the future in Machine Learning Engineer. Cheers!

    • @mrdbourke
      @mrdbourke День тому

      You’ve got this Fazri! Keep matching learning legend!

  • @iworeushankaonce
    @iworeushankaonce День тому

    @12:38 I expected rick roll

  • @leoaung3300
    @leoaung3300 День тому

    Pizza dad in real time before GTA 6!! Awesome vid

    • @mrdbourke
      @mrdbourke День тому

      😂 before GTA 6 hahaha

  • @rohitmondal4143
    @rohitmondal4143 День тому

    Hi, please bring in the optimised generation video of llm please

  • @mirabirhossain1842
    @mirabirhossain1842 День тому

    This video made me jealous. I am learning machine learning for like 3 years, hustle a bit, did some online unpaid projects, eventually got into a job, but in my job we only uses pretrained models api for various work. man i need to go out and train models. I am missing all the struggle and fun.

  • @holycake123
    @holycake123 День тому

    Which one is better - a MacBook Pro with the M3 Max chip, or a setup with an RTX 4090 for tasks like creating embeddings and building an RAG system?

  • @user-uz9qy1mf7x
    @user-uz9qy1mf7x 2 дні тому

    Why not start a subnet on Bittensor and get paid by network emissions for your work? you should really look into it

  • @semtex6412
    @semtex6412 2 дні тому

    "Machine Learning Engineer (at a *small* startup)" - meaning your "own" startup yea?

    • @mrdbourke
      @mrdbourke День тому

      Yes, this is our own startup. My brother Josh (in the video) and I are the co-founders.

  • @dhualshammaa2062
    @dhualshammaa2062 2 дні тому

    Do you have video showing how to great an phone app using LLM from A to Z ?

  • @user-kp6ys7yg7i
    @user-kp6ys7yg7i 2 дні тому

    Can we connect on social media to discuss project

  • @bonzita23
    @bonzita23 2 дні тому

    Do you raise the Pytorch flag every morning and sing the optimization loop song?

    • @mrdbourke
      @mrdbourke День тому

      Yes I do. I say good morning to my girlfriend and then I can’t help but singing “for an epoch in range… do the forward pass…” 🔥🎤🕺

  • @fingolete
    @fingolete 2 дні тому

    🥰 Dude. Im enjoying every minute of your work. Thank you, thank you, thank you.

    • @mrdbourke
      @mrdbourke День тому

      Glad you’re enjoying my friend! I really appreciate at the kind words.

  • @melalastro
    @melalastro 2 дні тому

    what are you expectations when chatgpt does this for free?

    • @mrdbourke
      @mrdbourke День тому

      ChatGPT already does a lot of it for free! You can take photos and upload them to ChatGPT and get information back. But ChatGPT doesn’t save your history, give insights into your trends or allow you to set macronutrient goals (maybe one day it will, who knows). Our models are also much faster than ChatGPT + work offline :D

  • @jag353
    @jag353 2 дні тому

    This was the best one so far man, appreciate it!

  • @mohammedissa9354
    @mohammedissa9354 2 дні тому

    one of the few Videos that will change your learning Journey. Thank You !