🎓 PhD candidate in Deep Learning for Remote Sensing data at UGR 🛰️
Granada, Spain
Joined October 2021
- Tweets889
- Following333
- Followers678
- Likes759
Jose Rodríguez retweeted
i understand you guys like lists
LeNet (1989)
LSTM (1997)
Deep Belief Networks (2006)
AlexNet (2012)
Word2Vec (2013)
GAN (2014)
VGG16/19 (2014)
InceptionNet (2014)
VAE (2014)
ResNet (2015)
U-Net (2015)
YOLO v1 (2015)
FastRCNN (2015)
FasterRCNN (2015)
Neural Style Transfer (2015)
Original Diffusion Models (2015)
SSD (2016)
DCGAN (2016)
AlphaGo(2016)
SqueezeNet (2016)
Transformer (2017)
CycleGAN (2017)
MobileNet (2017)
DenseNet (2017)
Mask R-CNN (2017)
BERT (2018)
YOLOv3 (2018)
BigGAN (2018)
StyleGAN (2018)
GPT-2 (2019)
DistilBERT (2019)
RoBERTa (2019)
XLNet (2019)
StyleGAN2 (2019)
DDPM (2020)
GPT-3 (2020)
DETR (2020)
DeiT (2020)
DALL-E (2020)
Alphafold 2
T5 (2020)
AlphaFold(2020)
GLIDE (2021)
CLIP (2021)
ViT (2021)
Swin Transformer (2021)
StyleGAN3 (2021)
Imagen (2022)
Flow Matching (2022)
Stable Diffusion (2022)
YOLOv7 (2022)
PaLM (2022)
GPT-4 (2023)
DINO v2 (2023)
SDXL (2023)
SAM (2023)
Alphafold 3 (2024)
Jose Rodríguez retweeted
I have been training Deep Learning models for 4 years now.
⚡ I can say that @LightningAI has been the most transformative library to improve my Deep Learning workflow. ⚡
Here are 5 benefits of using it:
Jose Rodríguez retweeted
Understanding Deep Learning
Impressive new book on understanding deep learning concepts.
Topics include fundamental building blocks, Transformers, GNNs, RL, diffusion models, and more.
Probably one of the most comprehensive and up-to-date overviews of deep learning that exist today.
Jose Rodríguez retweeted
The Transformer is a magnificient neural network architecture because it is a general-purpose differentiable computer. It is simultaneously:
1) expressive (in the forward pass)
2) optimizable (via backpropagation+gradient descent)
3) efficient (high parallelism compute graph)
Jose Rodríguez retweeted
Every year I teach CS theory, I find that I need to apologize less and less about having so much math content. CS practice is getting more & more mathy.
Now I just ask students if they want to their future job to be one where they read texts of first type, or the second type:
Jose Rodríguez retweeted
First Principles of Computer Vision, Columbia University
Really nice lectures on the physical and mathematical foundations of computer vision.
140 videos that you can watch at your pace. Slides are also provided to follow along.
Jose Rodríguez retweeted
most common neural net mistakes: 1) you didn't try to overfit a single batch first. 2) you forgot to toggle train/eval mode for the net. 3) you forgot to .zero_grad() (in pytorch) before .backward(). 4) you passed softmaxed outputs to a loss that expects raw logits. ; others? :)
Jose Rodríguez retweeted
Research Engineers are the people training and tuning state-of-the-art models like GPT-3, DALL-E, Imagen, Alphafold, etc
I spent 100s of hours coaching and mentoring junior and mid-career REs.
Here’s how you become an RE at a top-tier institution (Google, Meta, OpenAI,…) 🧵
Jose Rodríguez retweeted
One of the godfathers of AI @ylecu thinks AI is heading in the wrong direction. Here's what he thinks we should be doing instead to build machines that are as intelligent as humans. W/the great @strwbilly technologyreview.com/2022/06…
Jose Rodríguez retweeted
Read about new developments in deep learning with authors and researchers Daniel A. Roberts (@danintheory), Sho Yaida (@Shoyaida) and Boris Hanin (@BorisHanin) in their book The Principles of Deep Learning Theory: An Effective Theory Approach to Understanding Neural Networks. 👇
Jose Rodríguez retweeted
We’re excited to announce support for GPU-accelerated PyTorch training on Mac! Now you can take advantage of Apple silicon GPUs to perform ML workflows like prototyping and fine-tuning. Learn more: pytorch.org/blog/introducing…
Jose Rodríguez retweeted
Oriented bounding boxes are suitable for remote sensing imagery when you want orientation and more accurate measurements of objects size and aspect ratio 🛰
Our newest post and video by @misskellymore taking a look at YOLOv5-OBB.
Post: blog.roboflow.com/yolov5-for…
Video: youtube.com/watch?v=iRkCNo9-…
Jose Rodríguez retweeted
How to network at a conference?
Many students will be attending their first-ever *in-person* conference this year. How exciting! 🤩
Some tips on making the best of attending a conference. 🧵
Jose Rodríguez retweeted
Here it is: the first Learning on Graphs Conference! 🎊
We think this new venue will be valuable for the Graph/Geometric Machine Learning community.
What makes it so important+unique? See our blog post!
michael-bronstein.medium.com…
1/6
Jose Rodríguez retweeted
Contrastive Self-Supervised Learning can fall victim to dimensional collapse.
One reason I prefer non-contrastive methods these days.
Our ICLR'22 paper (openreview.net/forum?id=Yevs…) studies the phenomenon of dimensional collapsing in contrastive learning, in which the learned representation has a few dimensions collapsed to zero. Thanks all co-authors (@jingli9111, Pascal Vincent and @ylecun) for the great work! 1/
Jose Rodríguez retweeted
All ML projects which turned into a disaster in my career have a single common point:
🚨 I didn't understand the business context first, got over-excited about the tech, and jumped into coding too early.
Jose Rodríguez retweeted
Check out this interesting talk by @ylecun about future ideas for autonomous AI 🌟
How machines could learn like animals and human, how to learn models of the world and many more super exiting topics about intelligence 🧠
youtube.com/watch?v=DokLw1tI…
Jose Rodríguez retweeted
An official Ukrainian govt website with several good channels to support Ukraine & Ukrainians: donations, humanitarian aid, hosting refugees, becoming a medical volunteer, covering, fact-checking, and sharing news,...
war.ukraine.ua/support-ukrai…