Skip to content
View philip-bl's full-sized avatar

Organizations

@lesswrong-ru

Block or report philip-bl

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
philip-bl/README.md

Hi there. I'm Philip. Would you like to look at some code that I write? In that case you're in the right place. The table below showcases some projects I have written or contributed to.

Title Description
MNIST experiment from Tensorizing neural networks (Novikov et al. 2015) This is a replication of the MNIST experiment from Tensorizing Neural Networks. The author achieves a good result on MNIST without convolutional layers. The implementation of matrix tensor train layer in this repo is not very reusable, but it’s very simple to understand!
Deep Convolutional Tensor Network This repository contains the experiments I used in my paper Deep Convolutional Tensor Network presented at 2020 NeurIPS Quantum Tensor Networks in Machine Learning Workshop. In the paper, I study a novel hybrid neural network / tensor network machine learning model.
GPT-2 Chatbot This was a student group project at Skoltech. We fine-tuned GPT-2 on dialog corpora (this was before ChatGPT), built a chatbot and launched it on Telegram.
Tensor Train Decomposition for Recurrent Neural Networks This was a student group project at Skoltech. I replicated an experiment from Tensor-Train Recurrent Neural Networks for Video Classification (while accidentally getting SOTA results). My contributed code, Report describing the experiments
Bayesian Tensorized Neural Networks with Automatic Rank Selection Another student group project where we replicated a paper but failed to obtain the same results.
libcrap A small python library with a bunch of useful code written by me and used in many of my projects.

Pinned Loading

  1. tensorizing_neural_networks-novikov_2015 tensorizing_neural_networks-novikov_2015 Public

    MNIST experiment from Tensorizing neural networks (Novikov et al. 2015)

    Python 13 1

  2. dctn dctn Public

    Deep convolutional tensor network

    HTML 11 2

  3. tensor_train_rnn tensor_train_rnn Public

    Replace some weights matrices of linear layers of RNNs with matrices in tensor train format. Get better generalization and a lot less parameters.

    Jupyter Notebook 8 3

  4. gpt2_chatbot gpt2_chatbot Public

    Chatbot using GPT-2

    Jupyter Notebook 3 1

  5. vicontek/lrbtnn vicontek/lrbtnn Public

    Jupyter Notebook 2