Sunith Shetty

I Help IT Developers To Become Thought Leaders In Developer Communities Within 12 Months By Writing An Authoritative Book

[Santiago @svpino]

Here is the Amazon link: https://t.co/RrDdDBiMxV

[Santiago @svpino]

How to become GOD-LEVEL with Large Language Models.

Here are 50 hands-on projects with solutions that will teach you how Large Language Models work.

You don't need to solve all 50, but if you do, you'll be at the top 0.01% of the field.

It's all Python + Pytorch + SciKit-Learn + Pandas + Numpy + Matplotlib + Seaborn.

Here are the 50 problems from the book (link below):

Tokenization

1. Three tokenization schemes
2. Book lengths in characters, words, and tokens
3. Pandas frequency tables of token lengths
4. Token lengths in characters and bytes
5. Is tokenization compression?
6. Tokenization and compression in different languages
7. Translating between tokenizers

Embeddings

8. Distribution of cosine similarities
9. Sequential cosine similarity
10. Sequential number cosine similarity
11. Network graphs of cosine similarities
12. RSA to compare GPT-2 & BERT embeddings
13. Word similarity via distance and cosine
14. Linear semantic axes
15. Analogy vectors

Output logits

16. Softmax probability distributions
17. Probabilistic token selection
18. Token prediction accuracy
19. LLM loss function
20. Perplexity over sequences, texts, and models
21. Predict token position with linear and logistic regressions
22. Evaluating models with HellaSwag
23. Measuring language biases

Transformer outputs

24. Cosine similarities within and across layers
25. Category selectivity via cosine similarity
26. Current layer = previous layer + adjustments
27. Impact of layer-specific noise and scaling
28. Effective dimensionality of hidden layers
29. Hidden state dimensionality reduction
30. Sentiment analysis with decision trees
31. Logit lens
32. Patching hidden states in indirect object identification

Attention

33. QKV weights characteristics
34. QKV activation characteristics
35. Raw and softmax attention scores
36. Characteristics of attention adjustment magnitudes
37. Token prediction and attention KL divergences
38. Laminar profile of RSA and category selectivity
39. Token frequency, attention adjustments, QK^T
40. Downstream impacts of head silencing
41. Patching heads in IOI

MLP

42. MLP weights and activations characteristics
43. Characterizing the MLP progression
44. Grammar tuning in MLP projections
45. Minkowski distance, mutual information, and token positions
46. Statistics-based lesioning in MLP neurons
47. Supervised probing with XGBoost
48. "Can" vs. "can't" classification via logistic regression
49. Successive median-replacement of MLP activations
50. Recommender systems with MLP projections

Book link below.

[Sunith Shetty @sunithshetty]

@data4sci Much-needed live workshop. Thanks for putting this together!

[Tivadar Danka @TivadarDanka]

I’m teaching a live 4-hour workshop about the Mathematics of Machine Learning on January 24th, organized by Packt.

If you want to grab a ticket, here's a 40% discount code exclusive for my followers (valid until January 11th):

https://t.co/LyDD77X4uZ

[PyQuant News 🐍 @pyquantnews]

Algorithmic trading with Python AMA https://t.co/xcUZmpViiu

[Packt Data Science & Machine Learning @PacktDataML]

Isn’t it great to be part of a like-minded #data community to share ideas and #insights? We're here to do just that! Stay tuned for industry updates, info on production-ready tools, and exclusive access to vetted content in this thread. Let’s grow together! 🌱 #DataScience #ML

[Jeff Barr ☁️ @jeffbarr]

Natural Language Processing with #AWS AI Services - https://t.co/v3M0rlz060

"Derive strategic insights from unstructured data with Amazon Textract and Amazon Comprehend."

Congrats to my colleagues Mona & Prem for getting this to the finish line!

[Streamlit @streamlit]

🎈 Check out this new Streamlit book!! Congrats to Streamlit creator @tylerjrichards for writing "Getting Started with Streamlit for Data Science" - released today! 🥳

🤩 Learn more: https://t.co/2w0XQoZh3U
📕 Book: https://t.co/2CjdxFvbG0

#DataScience #Python #MachineLearning

[Sunith Shetty @sunithshetty]

@fchollet Many congratulations :)

[Sunith Shetty @sunithshetty]

@PacktPub @sg4women @RevistaSG @smasis @tarik_makota @maguirebrian @rajchakru @Somanath29

[François Chollet @fchollet]

Got this very nice book in the mail. All TensorFlow/Keras, with very readable code examples. Includes a section on StyleGAN, which will come in handy since I was trying to implement it the other day

[Saurabh Meher @SaurabhMeher]

@CMOMaharashtra
जय महाराष्ट्र,
जुहूतारा येथे ४२ पासून आमची हि पाचवी पिढी राहत आहे. आमचे घर आज BMC ३५१ तोडणार आहे. यात बिल्डरचा समावेश आहे. आमच्याकडे १९४२ पासूनचे सर्व कागदपत्रे आहेत. घराची १९९० दुरुस्त परवानगी आहे. आज बाळा साहेब ठाकरे असते तर ही वेळ आली नसती. आपले लक्ष असो.

[“the Phoeron” Colin J.E. Lupton @thephoeron]

Excited to announce our upcoming title, “Quantum Machine Learning for Azure Quantum” by myself and @michaelharries, a @PacktPub hands-on guide to real-world, cloud-native QML development. In memory of @peter_wittek all author proceeds will sponsor 3rd party open-source software.

[Tarek Amr 🍄 @gr33ndata]

Hands-On Machine Learning with Scikit-Learn and Scientific Python Toolkits (my new book) is ready for pre-ordering now! https://t.co/PZfqyH5U3N

[Prof Sara Lodge @LearNonsense]

Twitter, I have a request. It’s my daughter, Rosie’s 5th birthday tomorrow. But my partner has a fever, so we have to self-quarantine for 2 weeks. Her party is cancelled and grandparents barred. Could you send her some pics of your dogs (or animals generally) to cheer her up?

[Sunith Shetty @sunithshetty]

All cricket pundits, today you have no one to blame.

Everyone's strike rate looks like they are playing ODI/Test. But yes you have no Dhoni to blame today.

#INDvSA

[Anindh Kansal •Yoga & Meditation @anindhkansal]

#TrafficFines #TrafficRules
Some Countries Vs #India

[LF Decentralized Trust @lfdecentralized]

Thrilled to have another project join the Hyperledger greenhouse! Please welcome Hyperledger Besu, a Java-based #ethereum client formerly known as Pantheon. 1st Hyperledger project that can operate on a public blockchain. Read the blog by @PegaSysEng: https://t.co/Zclcy34S84

[Adam Parkhomenko @AdamParkhomenko]

brutal https://t.co/qsN5vI6JNr