This media is not supported in your browser
VIEW IN TELEGRAM
What goes on in the mind of BERT? Using an interactive visualization tool, we uncovered some surprisingly intuitive patterns.
👍1
Article is here: https://towardsdatascience.com/deconstructing-bert-distilling-6-patterns-from-100-million-parameters-b49113672f77?fbclid=IwAR35FclDpV-PqlWyfPKFqbXuy8YEpdNtuXGTKTl5bBnNKymeM0Ak9p4I-pk
Google Colab code is here: https://colab.research.google.com/drive/1vlOJ1lhdujVjfH857hvYKIdKPTD9Kid8?fbclid=IwAR2bQZMqXXXCYE3z9FCdRyKwfMzs4XVsF70fGkwMMFsIl0ClGGecN_BJ-gs
Google Colab code is here: https://colab.research.google.com/drive/1vlOJ1lhdujVjfH857hvYKIdKPTD9Kid8?fbclid=IwAR2bQZMqXXXCYE3z9FCdRyKwfMzs4XVsF70fGkwMMFsIl0ClGGecN_BJ-gs
Medium
Deconstructing BERT: Distilling 6 Patterns from 100 Million Parameters
From BERT’s tangled web of attention, some intuitive patterns emerge.
Media is too big
VIEW IN TELEGRAM
This picture is made of 195 billion pixels and all of the pixels are incredible and clear detail. Typical phone camera snaps 12-megapixel photos(12 million pixel). Just another advance of Computer Science. It also opens new challenges and research areas in the fields of #AI, #ML, #DL, #CV.
Thanks for Seeker and BigPixel Studios for creat video
Thanks for Seeker and BigPixel Studios for creat video
Must-read tutorial to learn sequence modeling from 'Analytics Vidhya' team and detailed explanation by real examples
The-Ultimate-Learning-Path-for-deep-learning.jpg
1.3 MB
Ultimate learning path of Deep Learning for 2019 in infographics. Image credited from Analytics Vidhya
This is a super cool resource: Papers With Code now includes 950+ ML tasks, 500+ evaluation tables (including SOTA results) and 8500+ papers with code. Probably the largest collection of NLP tasks I've seen including 140+ tasks and 100 datasets.
👍1
What Kagglers are mostly using for Text Classification?
👍1
imractical python.jpg
95.2 KB
Impractical PythonProjects by Lee Vaughan 2018
Github repository:
https://github.com/rlvaugh/Impractical_Python_Projects
https://github.com/rlvaugh/Impractical_Python_Projects
GitHub
GitHub - rlvaugh/Impractical_Python_Projects: Code & supporting files for chapters in book
Code & supporting files for chapters in book. Contribute to rlvaugh/Impractical_Python_Projects development by creating an account on GitHub.
Deep Learning Drizzle
Drench yourself in Deep Learning, Reinforcement Learning, Machine Learning, Computer Vision, and NLP by learning from these selected and exciting lectures!!
GitHub by Marimuthu Kalimuthu
Drench yourself in Deep Learning, Reinforcement Learning, Machine Learning, Computer Vision, and NLP by learning from these selected and exciting lectures!!
GitHub by Marimuthu Kalimuthu
You are deep learning enthusiast and Covolutions are unseperable part of your projects. In this tutorial given comprehensive guideline all about convolutions:
-> Convolution v.s. Cross-correlation
-> Convolution in Deep Learning (single channel version, multi-channel version)
-> 3D Convolution
-> 1 x 1 Convolution
-> Convolution Arithmetic
-> Transposed Convolution (Deconvolution, checkerboard artifacts)
-> Dilated Convolution (Atrous Convolution)
-> Separable Convolution (Spatially Separable Convolution, Depthwise Convolution)
-> Flattened Convolution
-> Grouped Convolution
-> Shuffled Grouped Convolution
-> Pointwise Grouped Convolution
-> Convolution v.s. Cross-correlation
-> Convolution in Deep Learning (single channel version, multi-channel version)
-> 3D Convolution
-> 1 x 1 Convolution
-> Convolution Arithmetic
-> Transposed Convolution (Deconvolution, checkerboard artifacts)
-> Dilated Convolution (Atrous Convolution)
-> Separable Convolution (Spatially Separable Convolution, Depthwise Convolution)
-> Flattened Convolution
-> Grouped Convolution
-> Shuffled Grouped Convolution
-> Pointwise Grouped Convolution