tabacof

7 ms·

1.
Slower Feels Smarter? Experimenting with AI Agent Latency(fin.ai)
1 points by tabacof 1y ago|0 comments
2.
How I lost 1000€ betting on CS:GO
1 points by tabacof 2y ago|0 comments
3.
I lost 1000€ betting on CS:GO – Foundations(tabacof.github.io)
2 points by tabacof 3y ago|1 comments
4.
by tabacof 3y ago
This post covers the foundations of e-sports betting with machine learning: financial decision-making, expected profits of a bet, multiple bets with the Kelly criterion, probability calibration, and the winner’s curse.
5.
by tabacof 3y ago
I started this blog recently, though some posts were written a few years ago (one even reached HN front page!): https://tabacof.github.io/
6.
Real-Time ML Models with Serverless AWS(tabacof.github.io)
1 points by tabacof 3y ago|1 comments
7.
by tabacof 3y ago
I wrote the blog post above based on the learnings I had when creating a MLOps solution using AWS Lambda for machine learning model deployment. Tl;dr: The post shows how to deploy a scikit-learn model pickle using Lambda and how to create a
8.
Name classification with ChatGPT: How does it compare to ML language models?(tabacof.github.io)
1 points by tabacof 4y ago|1 comments
9.
by tabacof 4y ago
In this blog post, I explore the problem of name classification with ChatGPT and 3 ML models of increasing complexity (logistic regression, FastAI LSTM, and HuggingFace DistilBERT). ChatGPT delivers the best accuracy of them all with no mod
10.
by tabacof 8y ago
Hi Soneca, I am also a Brazilian starting two very modest side business intending to sell to international customers: 1) https://www.deploir.com 2) https://www.whoworth.com What a coincidence, huh? In the beginning I
11.
by tabacof 9y ago
>Spark is a terribly inefficient solution to any known/stable data processing or analytics jobs. Can you expand on that?
12.
by tabacof 9y ago
Latin America is very unsafe. The list of top 50 cities by murder rate is very telling: https://en.wikipedia.org/wiki/List_of_cities_by_murder_rate I have lived most of my life in Brazil, and even though I am very priv
13.
by tabacof 11y ago
This is insane, this judge must be on a power trip. WhatsApp is now part of Brazilian social life and economy. Everyone here is part of many groups of friends, family or workmates, that is where most instant communication happens. In my com
14.
by tabacof 11y ago
Causality is a bit harder to integrate with current machine learning models as it's hard even with standard probabilistic graphical models. On the other hand, there has been a lot of work integrating deep neural networks with probabili
15.
by tabacof 12y ago
Torch7 has amazing GPU support. If you wanna send a tensor (array) to the GPU, you just type array:cuda(). All operations you do from now on will be on the GPU. More involved example using a neural network: http://code.cogbits.co