Skip to content
@tinkoff-ai

Tinkoff.AI

Tinkoff AI Center
README.md

Open source проекты Тинькофф

Hello! You are on the page of the Tinkoff AI Team. Here we will share our expertise and Open Source products.

Here you can find our libraries:

  • ETNA - to streamline data processing and avoid using multiple programs, we created ETNA - an open tool for analytics and time series forecasting
  • CORL - a library for RL research. The set is straightforward: one file - one algorithm. It can be adapted to any task that requires automated decision-making

We also provide the source code for our articles:

To learn more about how we work and develop the community, you can visit our page.

Where else to follow us:

  • Announcements of our events, useful links, and сurated set of memes in our Telegram channel
  • Telegram channel with job openings in various fields and streams
  • Telegram channel with Tinkoff IT digest

Pinned

  1. open-tlab open-tlab Public archive

    Примеры пропозалов для подачи заявки в Open.TLab

    24 3

  2. etna etna Public archive

    ETNA – Time-Series Library

    Python 841 77

  3. CORL CORL Public archive

    High-quality single-file implementations of SOTA Offline and Offline-to-Online RL algorithms: AWAC, BC, CQL, DT, EDAC, IQL, SAC-N, TD3+BC, LB-SAC, SPOT, Cal-QL, ReBRAC

    Python 981 112

  4. palbert palbert Public archive

    Code for the paper "PALBERT: Teaching ALBERT to Ponder", NeurIPS 2022 Spotlight

    Python 37 1

  5. eop eop Public archive

    Code for the paper "Showing Your Offline Reinforcement Learning Work: Online Evaluation Budget Matters", ICML 2022

    Jupyter Notebook 28 2

  6. sigir-2021 sigir-2021 Public archive

    4th place solution for the SIGIR 2021 challenge.

    Python 4

Repositories

Showing 10 of 19 repositories
  • etna Public archive

    ETNA – Time-Series Library

    Python 841 Apache-2.0 77 42 (1 issue needs help) 7 Updated Aug 9, 2023
  • ReBRAC Public archive

    Author's implementation of ReBRAC, a minimalist improvement upon TD3+BC

    Jupyter Notebook 50 Apache-2.0 6 0 0 Updated Aug 3, 2023
  • CORL Public archive

    High-quality single-file implementations of SOTA Offline and Offline-to-Online RL algorithms: AWAC, BC, CQL, DT, EDAC, IQL, SAC-N, TD3+BC, LB-SAC, SPOT, Cal-QL, ReBRAC

    Python 981 Apache-2.0 112 2 2 Updated Aug 3, 2023
  • katakomba Public archive

    Data-Driven NetHack Tools: Datasets (30+) and recurrent-baselines (AWAC, BC, CQL, IQL, REM)

    Python 63 3 0 0 Updated Jun 23, 2023
  • .github Public archive
    0 0 0 0 Updated May 31, 2023
  • palbert Public archive

    Code for the paper "PALBERT: Teaching ALBERT to Ponder", NeurIPS 2022 Spotlight

    Python 37 Apache-2.0 1 0 0 Updated Apr 8, 2023
  • d4rl Public archive Forked from Farama-Foundation/D4RL

    A benchmark for offline reinforcement learning.

    Python 1 Apache-2.0 265 0 0 Updated Mar 7, 2023
  • lb-sac Public archive

    Official implementation for "Q-Ensemble for Offline RL: Don't Scale the Ensemble, Scale the Batch Size", NeurIPS 2022, Offline RL Workshop

    Python 18 Apache-2.0 1 0 0 Updated Feb 27, 2023
  • sac-rnd Public archive

    Official implementation for "Anti-Exploration by Random Network Distillation", ICML 2023

    Python 47 Apache-2.0 4 0 0 Updated Feb 3, 2023
  • cnf Public archive

    Official implementation for "Let Offline RL Flow: Training Conservative Agents in the Latent Space of Normalizing Flows", NeurIPS 2022, Offline RL Workshop

    Python 12 Apache-2.0 0 0 0 Updated Jan 31, 2023
View all repositories