Latex quirks

its interesting how many quirks Latex has.

• I had to use a particular .sty  file for the paper.
  • to my dismay, I found its not sufficient to simple put it in the same folder as the .tex file
  • I had to do the following :
    • create directory where my other package files were present
    • place the .sty file there
    • do texhash
    • https://crmda.dept.ku.edu/guides/32.latex_config/32.latex_config.pdf
    • for me it was here : C:\Users\agoswami\AppData\Local\Programs\MiKTeX 2.9\tex\latex
• Bibliography
  • currently i am doing the following steps. i wonder if this can just be automated without using the command line each time i have to add a citation.
    • https://tex.stackexchange.com/questions/43325/citations-not-showing-up-in-text-and-bibiography
  • pdflatex main_trials
  • bibtex main_trials
  • pdflatex main_trials
  • pdflatex main_trials

Advertisements

System Design Interview

The system design interview can be of two kinds:

Object Oriented / Data Model :

• [Course] Programming-foundations-object-oriented-design
• http://www.andiamogo.com/S-OOD.pdf
• https://www.udemy.com/design-patterns-csharp-dotnet/

 

System Design :

• System Design Interview Questions

• Grokking the System Design Interview
• System Design

 

References:

• https://www.quora.com/How-do-I-answer-design-related-questions-like-design-a-parking-lot-in-an-Amazon-interview-Do-we-need-to-write-code-for-design-questions
• https://www.quora.com/How-do-I-prepare-to-answer-design-questions-in-a-technical-interview

Things you do as a well rounded ML Engineer / Data Scientist

I came across this list somewhere on the web.

I felt the list nicely captures the different kinds of skills a well rounded ML Engineer/Data Scientist develops over his/her career:

Architect data analytics solutions Access and Design Data Warehouse / DSS infrastructures Build and maintain data-driven applications Conduct AI / ML related research Conduct statistical analysis of the data Create BI reports and dashboards Create data visualizations Create / update predictive models using machine learning Develop AI / ML tools and libraries for internal or external customers Develop new data pipelines (ETL / ELT) Integrate new data sources for my organization Maintain and monitor data pipelines Maintain and monitor data warehouse Make decisions based on data-driven insights Operationalize machine learning models Support the AI / ML education

Python dictionary == operator

Its interesting to note that python dictionaries support the == operator

References:

• https://stackoverflow.com/a/5635309
• https://stackoverflow.com/questions/17217225/what-does-the-operator-actually-do-on-a-python-dictionary
• https://www.laurentluce.com/posts/python-dictionary-implementation/
  • This is a nice read on Python’s dictionary implementation using hash tables
• https://hg.python.org/cpython/file/6f535c725b27/Objects/dictobject.c#l1839

TensorFlow – save, restore, freeze models

It can be quite tricky getting around to deploying TF models –

In fact there are multiple ways to save/load TF models, with each serving a slightly different purpose / use-case.

• simple_save
• saver
• estimator
• keras
• tflite

 

References:

• https://blog.metaflow.fr/tensorflow-how-to-freeze-a-model-and-serve-it-with-a-python-api-d4f3596b3adc
• https://gist.github.com/omimo/5d393ed5b64d2ca0c591e4da04af6009

Code:

• TensorFlow8_cifar.py
• TensorFlow8_cifar_load.py
• freeze.py

 

Consistent Hashing

References:

• https://stackoverflow.com/questions/144360/simple-basic-explanation-of-a-distributed-hash-table-dht
• http://michaelnielsen.org/blog/consistent-hashing/
• http://michaelnielsen.org/blog/lecture-course-the-google-technology-stack/
  • this is wonderful reading on Google’s technology stack!

 

Interfaces in Python : DuckTyping + ABCs

References:

• http://masnun.rocks/2017/04/15/interfaces-in-python-protocols-and-abcs/

Code: