pavlovskyi

  Location: poland
  Remote:yes
  Willing to relocate:no
  Technologies:python pytorch deep learning, ml, dl
  Résumé/CV: huggingface.co./s3nh ; github.com/s3nh
  Email: [email protected]

Mostly building deep learning solutions with different data sources, now focusing on cv and llms. Additive value is a must, big fan of model quantization, 7+ years of exp in building models, want to improve my skills and earn more money. All best

Location: Poland Remote: yes Willing to relocate: Yes Technologies:Python, Pytorch, Deep Learning, Machine Learning, Spark, Gcp, APIs. Résumé/CV: Email: [email protected]

Working with ml/dl algorithm for 6+ years, now working as an principal data scientist. I want to deal with serious datasets and solve real case problems. You can look at some of my work on github.com/s3nh and huggingface.com/s3nh

Data Scientist with 6+ years of experience in building models on various of data sources. Now mainly focusing on computer vision and nlp task on relatively big amount of data. Fun of spare time projects and paperspace platform.

Location: Poland Remote: preferable Willing to relocate: Technologies: python, REST, pytorch, spark Résumé/CV: github.com/s3nh, huggingface.com/s3nh Email: [email protected]

About:

Data scientist with 6 years of experience with statistical modeling, machine learning deep learning. Interested in computer vision and use Ai for goods. Mainly focused on python/pytorch. Creating solutions from ideation to deployment phase. Technologies: Python/Pytorch, hdl sql rest apis, azure pipelines, github, docker etc. Interested in developing creative solutions, not interested in data engineering positions.

Location: Poland

Remote: Yes

Looking for: Full time

Visa: No

You can watch some of my work at: - Github.com/s3nh - s3nh.github.io

or contact me at: [email protected]

Feel free to reach me out, All best

Ehh, I totally forgot where I am, for a few moments.

I wont click but it sounds like best prepared ‚yo mama joke’

Improvement - which way to choose.

I like the idea, never heard of it before but from I always lived thinking that reducing variance between Slavic languages is needed because (as a Pole) I understand context but it is hard to perform conversation with different pairs (Slavian PL for example). Good to know that that project exist, maybe it is an niche to use some ml/ai methods to generate pretrained model for generating speech etc. Especially in huggingface like era it could be relatively easy and helpful.And also catchy, just look at that buzzwords ' AI help to unify slavic languages'.

I like the idea, never heard of it before but from I always lived thinking that reducing variance between Slavic languages is needed because (as a Pole) I understand context but it is hard to perform conversation with different pairs (Slavian PL for example). Good to know that that project exist, maybe it is an niche to use some ml/ai methods to generate pretrained model for generating speech etc. Especially in huggingface like era it could be relatively easy and helpful.And also catchy, just look at that buzzwords ' AI help to unify slavic languages'.

For a long time I lived thinking that I did something wrong working with nvidia drivers. Ubuntu installation comes with no problem every time I did it, but there was tremendous amount of issues which comes exact from installing nvidia drivers. It was so irritating and not intuitive that I always thought that this incosistency and issue generating flow is just matter of time. But it is still in game and installing CUDA Toolkit gave you that feeling like you are applying on junior software engineer to nvidia and it is a technical interview.

Awesome interpretation! I did not think about numbers interpretation for a long time, but numbers were always my passion, especcialy as a child. The fun fact is that I gave any number some kind of uninterpretable personality, some kind like information about its historical behaviour. and its allows me to like them more or not, give them positive or negative judging. So in case of addition, multiplication, etc (which I can do blazingly fast from my youngest) I see it as some kind of story in which numbers really meet each other and produce some results.

There is no other feeling in my mind which give me that amount of understading,but it is an understanding which cannot be formulate properly to other person. I feel it as an phenomena which origin started in my mind and grow there for my whole life (which is highly correlated with my introvert pov). Thanks for that thoughts, all best.

It could be really to generate synthetic data for semantic segmentation tasks.

Knowing the possibilities of sota AI algorithms and its limitation I would say that I would never replace any human with CEO if it comes to an process in which you have to make decisions.

The main issue with 'AI' is that it make an judgment by the data it was feed on through the training process. So there is no spark of creativity which will result in behaviours which are far away from expected value.

Unpopular opinion: I know that hype train change its direction and know we are stuffed with buzzwords like 'Ai algo solve mathematical problem', but the question is how the test dataset was prepared and where is the border between excellent memorization and average creativity. Because assuming that we have an infinite number of resources we can teach any algorithm literally anything in limited amount of time.

I observed that kind of experiments for a long time, it is catchy, relatively easy to implement and it gives that feeling that artificial intelligence create something by itself. But as long as it will stricly dependent of distribution of the data that we put in it, It will always be some kind of less or more visible modification of input data.

Yes, I totally agree that ‚why do we need this’ is crucial. But I ve seen to many cases in which main target was to develop instead pf give real benefit. Maybe it is caused by some companies immaturity in Ai field, just guessing.

But it is just an assumption. I work as a data scientist for 5+ years and from practical point of view, it is not just data wrangling. It is worth to mention that going through that logic we assume that programmer fully understand how to develop model in production and how to handle it in some border cases, which is not true.