I came across an article about the large scale Face Recognition system at @odnoklassniki.
— Vladimir Iglovikov (@viglovikov) July 20, 2020
I liked how they created the train set. The trick about filtering false positives is also smart.
Translated it into English.https://t.co/ifOunH470b#DeepLearning #ArtificialInteligence
A great primer on Bayesian Optimization: https://t.co/NJIG77N2Uq
— Martin Görner (@martin_gorner) July 17, 2020
Finally I know how AI Platform's hyperparameter tuning feature works. (https://t.co/A2jdwwSo5g)
Disney uses PyTorch for animated character recognition and to speed up its video processing pipeline. @DTCITechnology engineers also contributed new features to the Torchvision domain library. https://t.co/2DxwcTRtoF
— PyTorch (@PyTorch) July 16, 2020
Explaining RNNs without neural networks. A summary of what I learned building RNNs with just linear algebra, starting from @karpathy's code then reading fastai book from @jeremyphoward / @GuggerSylvain plus @iamtrask's grokking DL book. Thanks guys! https://t.co/6wrW3cFz8J
— Terence Parr (@the_antlr_guy) July 10, 2020
Synthetic Data for Deep Learning
— hardmaru (@hardmaru) July 3, 2020
Synthetic data is increasingly used for training large deep learning models, especially in computer vision. This is probably the most comprehensive survey (~ 100 pages) on leveraging synthetic data in various DL areas.https://t.co/sXZOwpFMA1 pic.twitter.com/jYgQE7vxsL
Offline reinforcement learning talk, which I prepared for a seminar at Google -- here is a recording I made (while rehearsing...). This talk covers CQL, D4RL, and briefly touches on AWR, AWAC, and some theoretical foundations.https://t.co/81agcxLqSa
— Sergey Levine (@svlevine) July 1, 2020
Model-based Reinforcement Learning: A Survey: https://t.co/5tvG5Npx4v
— Denny Britz (@dennybritz) July 1, 2020
“A systematic categorization of planning-learning integration, including where to start planning, what budgets to allocate, how to plan, and how to integrate planning in the learning and acting loop” pic.twitter.com/m1DEgdeiU8
Evaluation of Text Generation: A Survey: https://t.co/d666nSGT7Y
— Denny Britz (@dennybritz) June 29, 2020
This looks like a neat survey of human, automated and learned metrics. It also has one of the longest bibliography sections I’ve ever seen, 16 pages….
Corrupt, sparse, irregular and ugly: Deep learning on time serieshttps://t.co/xCU6aqFgzf
— Jeremy #Masks4All Howard (@jeremyphoward) June 24, 2020
Deep Learning Based Text Classification: A Comprehensive Review: https://t.co/jvgMqUAt9T
— Denny Britz (@dennybritz) June 24, 2020
A neat survey on text classification techniques and datasets, all the way from doc2vec to BERT. pic.twitter.com/CYzToUyNKl
I looked at 200 tools for developing & deploying machine learning applications:
— Chip Huyen (@chipro) June 23, 2020
- how the market evolved over time
- the difference between ML applications that traditional software engineering applications
- open-source vs open-core business modelhttps://t.co/LmAcKADRJK
Take a walk through the Life of a @numba_jit Kernel and experience the compilation pipeline that takes user-defined functions in #Python to #CUDA Kernels. https://t.co/SUoooTPoil
— RAPIDS AI (@rapidsai) June 16, 2020
