Spark in me - Internet, data science, math, deep learning, philo

snakers4 @ telegram, 1444 members, 1691 posts since 2016

All this - lost like tears in rain.

Data science, deep learning, sometimes a bit of philosophy and math. No bs.

Our website
- spark-in.me
Our chat
- goo.gl/WRm93d
DS courses review
- goo.gl/5VGU5A
- goo.gl/YzVUKf

snakers4 (Alexander), December 20, 12:12

Spell-checking on various scales in Russian

Bayes + n-gram rules = spell-checker for words / sentences

habr.com/company/joom/blog/433554/

#nlp

Исправляем опечатки в поисковых запросах

Наверное, любой сервис, на котором вообще есть поиск, рано или поздно приходит к потребности научиться исправлять ошибки в пользовательских запросах. Errare...


www.facebook.com/nipsfoundation/videos/203530960558001

Neural Information Processing Systems

Welcome to NeurIPS 2018 Turorial Sessions. This tutorial on Visualization for Machine Learning will provide an introduction to the landscape of ML visualizaions, organized by types of users and their...


snakers4 (Alexander), December 19, 08:16

DS/ML digest 32

Highlights:

- A way to replace softmax in NMT;

- Large visual reasoning dataset;

- PyText;

spark-in.me/post/2018_ds_ml_digest_32

#digest

#deep_learning

#data_science

2018 DS/ML digest 32

2018 DS/ML digest 32 Статьи автора - http://spark-in.me/author/snakers41 Блог - http://spark-in.me


snakers4 (Alexander), December 17, 13:18

ganbreeder.app/

A collaborative tool for discovering images.


snakers4 (Alexander), December 17, 09:24

PyText

- PyText github.com/facebookresearch/pytext from Facebook:

- TLDR - FastText meets PyTorch;

- Very similar to AllenNLP in nature;

- Will be useful if you can afford to write modules for their framework to solve 100 identical tasks (i.e. like Facebook with 200 languages);

- In itself - seems to be too high maintenance to use;

I will not use use it.

#nlp

#deep_learning

facebookresearch/pytext

A natural language modeling framework based on PyTorch - facebookresearch/pytext


snakers4 (Alexander), December 15, 17:50

www.youtube.com/watch?v=ZKQp28OqwNQ

BigGANs: AI-Based High-Fidelity Image Synthesis
This episode was supported by insilico.com. "Anything outside life extension is a complete waste of time". See their papers: - Papers: www.ncbi.nlm.n...

snakers4 (Alexander), December 14, 19:10

PyText?

NLP library build on top of PyTorch 1.0 by Facebook?

- No repo link though (github.com/facebookresearch/pytext)

- The paper also mentions the same limited API as AllenNLP has ... =(

research.fb.com/publications/pytext-a-seamless-path-from-nlp-research-to-production/

facebookresearch/pytext

A natural language modeling framework based on PyTorch - facebookresearch/pytext


snakers4 (Alexander), December 14, 03:57

youtu.be/kSLJriaOumA

A Style-Based Generator Architecture for Generative Adversarial Networks
Paper (PDF): stylegan.xyz/paper Authors: Tero Karras (NVIDIA) Samuli Laine (NVIDIA) Timo Aila (NVIDIA) Abstract: We propose an alternative generator a...

snakers4 (Alexander), December 10, 15:14

Forwarded from Админим с Буквой:

habr.com/post/432686/

WireGuard — прекрасный VPN будущего?

Наступило время, когда VPN уже не является каким-то экзотическим инструментом бородатых сисадминов. Задачи у пользователей разные, но факт в том, что VPN стал...


snakers4 (Alexander), December 10, 04:27

Simpsons paradox

Nice explanation

towardsdatascience.com/simpsons-paradox-and-interpreting-data-6a0443516765

#data_science

Simpson’s Paradox and Interpreting Data

The challenge of finding the right view through data


snakers4 (Alexander), December 09, 07:59

DS/ML digest 31

Highlights of the week:

- PyTorch 1.0 released;

- Drawing with GANs;

- BERT explained;

spark-in.me/post/2018_ds_ml_digest_31

#digest

#deep_learning

#data_science

2018 DS/ML digest 31

2018 DS/ML digest 31 Статьи автора - http://spark-in.me/author/snakers41 Блог - http://spark-in.me


snakers4 (Alexander), December 08, 05:17

PyTorch 1.0 release

View Release:

 github.com/pytorch/pytorch/releases/tag/v1.0.0

#deep_learning

pytorch/pytorch

Tensors and Dynamic neural networks in Python with strong GPU acceleration - pytorch/pytorch


snakers4 (Alexander), December 07, 03:06

youtu.be/AGm3hF_BlYM

This AI Learns Human Movement From Videos
The paper "Towards Learning a Realistic Rendering of Human Behavior" is available here: compvis.github.io/hbugen2018/ Pick up cool perks on our Patre...

snakers4 (Alexander), December 06, 09:04

Painting with GANs

This looks just awesome.

I guess it will not work in real resolutions yet.

gandissect.res.ibm.com/ganpaint.html?project=churchoutdoor&layer=layer4

#deep_learning

Painting with GANs from MIT-IBM Watson AI Lab

This demo lets you modify a selection of meaningful GAN units for a generated image by simply painting.


snakers4 (Alexander), December 05, 14:35

This kind of mirrors my own old post

spark-in.me/post/epistemic-responsibility

Эпистемологическая ответственность

Почему мы должны нести ответственность за идеи, в которые верим и которыми пользуемся Статьи автора - http://spark-in.me/author/snakers41 Блог - http://spark-in.me


Forwarded from Вастрик.Пынь:

💌 Вастрик.Инсайд #37: Этикет, грамотность и обжорство в информационном поле

Давно хотелось поговорить на тему осознанного потребления информации и организации личного инфополя. Почему самоограничения и разборчивость — хорошо, но удалять все аккаунты и уходить в лес — не очень хорошая крайность. Если вы сами не организуете своё инфополе, это сделают за вас.

В конце выпуска есть формочка для вопросов на итоговый выпуск. Воспользуйтесь.

vas3k.ru/inside/37/

Вастрик.Инсайд #37

Этикет, грамотность и обжорство в информационном поле


snakers4 (Alexander), December 02, 09:40

A cheeky ML/DS themed sticker pack for our channel

Thanks to @birdborn for his art.

You are welcome to use it:

t.me/addstickers/ML_spark_in_me_by_BB

If you would like to contribute / create your own stickers - please ask around in our channel chat.

#data_science

snakers4, November 30, 11:01

Channel Edit Photo

snakers4 (Alexander), November 29, 08:10

Article about the reality of CV in Russia / CIS

(RU)

cv-blog.ru/?p=253

Also a bit on how to handle various types of "customers", who want to contract CV systems from you.

Warning - too much harsh reality)

#deep_learning

snakers4 (Alexander), November 28, 11:55

DS/ML digest 30

spark-in.me/post/2018_ds_ml_digest_30

#digest

#deep_learning

#data_science

2018 DS/ML digest 30

2018 DS/ML digest 30 Статьи автора - http://spark-in.me/author/snakers41 Блог - http://spark-in.me


snakers4 (Alexander), November 27, 18:55

snakers4 (Alexander), November 25, 05:29

Creating a new user

With the above hack, user creation can be done as easy as:

USER="YOUR_USER" && \

GROUP="YOUR_GROUP" && \

sudo useradd $USER && \

sudo adduser $USER $GROUP && \

sudo mkdir -p /home/$USER/.ssh/ && \

sudo touch /home/$USER/.ssh/authorized_keys && \

sudo chown -R $USER:$USER /home/$USER/.ssh/ && \

sudo wget -O - github.com/$USER.keys | sudo tee -a /home/$USER/.ssh/authorized_keys#linux

snakers4 (Alexander), November 25, 04:59

Getting your public key from Github ... with wget!

I kind of saw it when installing Ubuntu 18 from scratch. But it is super awesome!

wget -O - github.com/snakers4.keys >> test

Just replace test with your authorized_keys file and profit!

#linux

snakers4 (Alexander), November 23, 12:43

Jupyter extensions

Looks like they are near end of their support.

Alas.

On a fresh build you will need this

conda install notebook=5.6

To use them.

Will need to invest some time into making Jupyter Lab actually usable.

#data_science

snakers4 (Alexander), November 23, 08:21

TDS article follow-up

TDS also accepted a reprint of the article

towardsdatascience.com/winning-a-cft-2018-spelling-correction-competition-b771d0c1b9f6

#nlp

Winning a CFT 2018 spelling correction competition

Or building a task-agnostic seq2seq pipeline on a challenging domain


snakers4 (Alexander), November 22, 11:56

Our victory in CFT-2018 competition

TLDR

- Multi-task learning + seq2seq models rule;

- The domain seems to be easy, but it is not;

- You can also build a pipeline based on manual features, but it will not be task agnostic;

- Loss weighting is crucial for such tasks;

- Transformer trains 10x longer;

spark-in.me/post/cft-spelling-2018

#nlp

#deep_learning

#data_science

Winning a CFT 2018 spelling correction competition

Building a task-agnostic seq2seq pipeline on a challenging domain Статьи автора - http://spark-in.me/author/snakers41 Блог - http://spark-in.me


snakers4 (Alexander), November 22, 11:49

Problems with GPUs in DL box

Cryptic messages like:

GPU is lost

Usually this is either:

- PSU;

- Or bad PCIE contact;

- Too much load on PCIE bus;

#deep_learning

snakers4 (Alexander), November 19, 13:59

Forwarded from Loss function porn:
"80 years of AI research. Epic battle between connectionist (~neural networks) and symbolic (~rule based) methods. Who will win?"
👤 @OriolVinyalsML (twitter)
📉 @loss_function_porn

snakers4 (Alexander), November 15, 08:09

DS/ML digest 29

spark-in.me/post/2018_ds_ml_digest_29

#digest

#deep_learning

#data_science

2018 DS/ML digest 29

2018 DS/ML digest 29 Статьи автора - http://spark-in.me/author/snakers41 Блог - http://spark-in.me


snakers4 (Alexander), November 14, 08:36

An intro to RL

Though published by OpenAI with TF, this is simply amazing:

- spinningup.openai.com/en/latest/spinningup/rl_intro.html

#rl

snakers4 (Alexander), November 12, 19:26

youtu.be/zL6ltnSKf9k

This AI Learned To Isolate Speech Signals
The paper "Looking to Listen at the Cocktail Party: A Speaker-Independent Audio-Visual Model for Speech Separation " is available here: looking-to-li...

snakers4 (Alexander), November 12, 09:09

When it is colder, under full load GPUs run at 70C