Numyard_DataScience

Data Science Group that focuses on high quality data science programs, discussions and problem-solving

[Numyard_DataScience @numyard_ds]

The easiest way to have your code available on the #Python path when using virtualenv and pip is to have a https://t.co/pLEphbwFf5 file and install your project in editable mode when developing

Credit: pytest-django plugin documentation https://t.co/Lpma782pia

#DataScience

[Numyard_DataScience @numyard_ds]

If you're learning AI/ML/DL - one of the hardest pieces is knowing what projects to try next!

This is a golden link!

Over 200 different tutorials across nearly every major branch of Machine Learning!!!

https://t.co/QgwKqF7GDR

[Numyard_DataScience @numyard_ds]

A great repository containing material that teaches you #Python and important topics like inheritance and debugging.

It emphasizes on handling data which makes it a nice starting point for someone jumping into data analysis with Python
https://t.co/LrS19tOSYZ

#DataScience

[Numyard_DataScience @numyard_ds]

If you have a widescreen monitor, this new Rstudio 1.4 feature may be relevant to your interests.

https://t.co/9snTbG4Rjr

#rstats #datascience #MachineLearning

[Numyard_DataScience @numyard_ds]

Here's a nice thread by Elvis, on course recommendations to get started with machine learning

https://t.co/Di2Kiv5tN4

#machinelearning #datascience

[Numyard_DataScience @numyard_ds]

Tokenizers 0.9 is out, and you are going to love it!
- Full support for Unigram (SentencePiece)
- Numpy arrays as input
- Many new components for better customization
- Shiny new docs: https://t.co/6o5a0up1tO

Check how to create a byte-level Unigram:

#deeplearning #nlp

[Numyard_DataScience @numyard_ds]

Zero Shot Classification with Huggingface Pipeline:

Video Link : https://t.co/Dp3UZAsHYI

#machinelearning #deeplearning #nlp #datascience

[Numyard_DataScience @numyard_ds]

Another Python standard library gem: difflib.get_close_matches returns a list of the best “good enough” matches.

#python #machinelearning #datascience

[Numyard_DataScience @numyard_ds]

This Keras guide on fine-tuning EfficientNets is the one of the best we have seen to date. Includes many practical tips.

https://t.co/2TBZfKo6sP

#deeplearning #machinelearning #datascience

[Numyard_DataScience @numyard_ds]

Fun time. Want to generate your daily horoscope?

AstroGPT - GPT-2 model fine-tuned on Western zodiac signs

Input today's date on:

https://t.co/EIwKPrRyFl

#nlp #machinelearning #deeplearning #datascience

[Numyard_DataScience @numyard_ds]

#fastpages is a backbone of #fastai library that brings advanced #Python constructs like typedispatch, store_attr(), delegates, @patch..

See examples below:

https://t.co/6IGfww4HXo

#machinelearning #deeplearning #datascience

[Numyard_DataScience @numyard_ds]

Here's a reinforcement learning from human feedback to train language models for summarization. The resulting models produce better summaries than 10x larger models trained only with supervised learning:

https://t.co/5tjOZdI395

#machinelearning #deeplearning #datascience

[Numyard_DataScience @numyard_ds]

If you found qlalchemy for basic DB querying confusing, here's a tiny library, `fastsql` by Jeremy Howard that makes it quite a bit simpler.

Here's an example of grabbing data from a DB.

https://t.co/yqBukXmU4I

#machinelearning #python #deeplearning #datascience

[Numyard_DataScience @numyard_ds]

This is 100% #Julia language code. It trains a dense neural network using the #Keras #Python library. If you replace the first 3 lines with "import tensorflow as tf; from #tensorflow import keras", you can run the exact same code in Python.

#machinelearning #datascience

[Numyard_DataScience @numyard_ds]

several ways to interactively display results of a calculation:

# Two liner
>>> q = 3.5 * 1.5
>>> q
5.25

# Second expression
>>> q = 3.5 * 1.5; q
5.25

# Walrus
>>> (q := 3.5 * 1.5)
5.25

# Invisible
>>> q = 3.5 * 1.5

#python #datascience #machinelearning

[Numyard_DataScience @numyard_ds]

Ultimate Python study guide for newcomers and professionals alike..

https://t.co/F83ptbToxB

#python #machinelearning #datascience

[Numyard_DataScience @numyard_ds]

This blog post provides another great example of how to easily build a semantic search engine using Huggingface transformers and Elasticsearch.

https://t.co/F83ptbToxB

#nlp #deeplearning #machinelearning #datascience

[Numyard_DataScience @numyard_ds]

This repo contains a few survey papers related to machine learning.

Survey papers are a great way to review current trends in ML and other related areas, and can help you find interesting research topics.

https://t.co/wdZhJpKOcy

#machinelearning #deeplearning #datascience

[Numyard_DataScience @numyard_ds]

E-BERT proposes ideas on how to improve a BERT based language model for e-commerce use cases.

The proposed method enhances BERT by modeling product & phrase-level knowledge enabling capabilities

https://t.co/q7vpNRWZc8

#machinelearning #deeplearning #nlp #datascience

[Numyard_DataScience @numyard_ds]

Here's an amazingly easy way to make custom colormaps:

custom_cmap = LinearSegmentedColormap.from_list(
"my_cmap",
["red", "black", "white"])

matplotlib docs are the gift that never stops giving.

https://t.co/ZFnTttwpT3

#python #machinelearning #datascience