Marton Trencseni - Sun 07 May 2023 • Tagged with gpt, python, gnews, ai
I show how I used GNews and the OpenAI API to build GPT News Poet, a fun toy site showing silly AI poems based on today's news.
Marton Trencseni - Wed 21 December 2022 • Tagged with similarity, python, word2vec, doc2vec, pytorch, tensorflow
I use Doc2Vec to try to find pairs of similar API functions between Pytorch and Tensorflow.
Marton Trencseni - Sat 10 December 2022 • Tagged with similarity, python, gensim, word2vec, doc2vec, pyml
In the previous posts, I used the Doc2Vec neural network architecture to compute the similarities between my blog posts, and explored the quality of the scores. In this final post, I show how I added the final Articles You May Like recommendation sections to the blog — it's live!
Marton Trencseni - Sun 04 December 2022 • Tagged with similarity, python, gensim, word2vec, doc2vec, pyml
In the previous post, I used the Doc2Vec neural network architecture to compute the similarities between my blog posts. In this second post I investigate the results further by examining clusters in graphs.
Marton Trencseni - Sat 03 December 2022 • Tagged with similarity, python, gensim, word2vec, doc2vec, pyml
One of the things I learned at Facebook is the power of recommendations. Examples are People You May Know (PYMK), Groups You May Like (GYML) and Pages You May Like (PYML). Inspired by these, I am planning to add an Articles You May Like widget to Bytepawn, based on the semantic similarity of blog posts. I use the Doc2Vec neural network architecture to compute the similarity between my blog posts, and return the top 3 recommendations for each page.
Marton Trencseni - Sat 22 January 2022 • Tagged with wordle, monte-carlo
I present a simple Monte Carlo solution which finds a 25-letter-unique Worlde wordlist in about 10 minutes. 
Marton Trencseni - Fri 21 January 2022 • Tagged with wordle, monte-carlo
I improve on the previous brute-force Monte Carlo approach for attacking the Wordle coverage problem. 
Marton Trencseni - Wed 19 January 2022 • Tagged with wordle, monte-carlo
I show a simple brute-force Monte Carlo approach for attacking the Wordle coverage problem. 
Marton Trencseni - Thu 06 May 2021 • Tagged with python
I describe a real world use-case where a simple, brute force search based solution worked really well, making more sophisticated Machine Learning unnecessary.
Marton Trencseni - Fri 09 April 2021 • Tagged with python, pytorch, cnn, torchvision, mnist, autoencoder
I measure how the classification accuracy of quantized Autoencoder neural network varies with encoding bits on MNIST digits.
Marton Trencseni - Sun 04 April 2021 • Tagged with python, pytorch, cnn, torchvision, mnist, autoencoder
I investigate how much information an Autoencoder neural network encodes for MNIST digits.
Marton Trencseni - Tue 02 March 2021 • Tagged with python, pytorch, torchvision, mnist, gan
I train a Pytorch Classic MNIST GAN on Google Colab to generate MNIST digits.
Marton Trencseni - Thu 12 December 2019 • Tagged with pytorch
2019 was another big year for Pytorch, one of the most popular Deep Learning libraries out there. Pytorch has become the de facto deep learning library used for research thanks to it’s dynamic graph model which allows fast model experimentation. It’s also become production ready, with support for mobile and infrastructure tooling such as Tensorboard.
Marton Trencseni - Thu 21 November 2019 • Tagged with machine, learning, fetchr, skl, calibration
I show calibration curves for four different binary classification Scikit-Learn models we built for delivery prediction at Fetchr, trained using real-world data: LogisticRegression, DecisionTree, RandomForest and GradientBoosting. 
Marton Trencseni - Thu 14 November 2019 • Tagged with python, pytorch, reinforcement, learning, openai, gym
I use simulated self-play by ranking episodes by summed reward. Game outcomes are divided in two by cutting at the median, winners are assigned +1 rewards, losers are assigned -1 rewards, like in games like Go and Chess. Unlike naive policy gradient descent used in previous posts, this version solves all OpenAI classic control problems, albeit slowly.
Marton Trencseni - Tue 12 November 2019 • Tagged with python, pytorch, reinforcement, learning, openai, gym
I try to generalize the policy gradient algorithm as introduced earlier to solve all the OpenAI classic control problems. It works for CartPole and Acrobot, but not for Pendulum and MountainCar environments.
Marton Trencseni - Tue 29 October 2019 • Tagged with machine, learning, fetchr, skl
Opportunities for automating, optimizing and enabling processes with ML at a delivery company such as Fetchr are plentiful. We put three families of ML models into production. These 3 areas are: Scheduling, Notifications and Operational choice.
Marton Trencseni - Tue 22 October 2019 • Tagged with python, pytorch, reinforcement, learning, openai, gym, cartpole
The CartPole problem is the Hello World of Reinforcement Learning, originally described in 1985 by Sutton et al. The environment is a pole balanced on a cart. CartPole is one of the environments in OpenAI Gym, so we don't have to code up the physics. Here I walk through a simple solution using Pytorch.
Marton Trencseni - Sun 25 August 2019 • Tagged with python, pytorch, cnn, go
Using historic gameplay between strong Go players as training data, a CNN model is built to predict good Go moves on a standard 19x19 Go board.
Marton Trencseni - Fri 02 August 2019 • Tagged with pytorch, skl, arabic, fetchr
While working on arabic-vs-rest classification, I was curious how good out-of-the-box models perform with publicly available data, and then compare that with what we can achieve with internal data / features derived from millions of deliveries. We train Scikit-learn and Pytorch models for this classification task and achieve 90% prediction accuracy on publicly available data and out-of-the-box models, while internally 99% is achievable.
