306K subscribers
4K photos
707 videos
17 files
4.58K links
Погружаемся в машинное обучение и Data Science

Показываем как запускать любые LLm на пальцах.

По всем вопросам - @haarrp

@itchannels_telegram -🔥best channels

Реестр РКН: clck.ru/3Fmqri
Download Telegram
Deep Image Compression using Decoder Side Information

Code: https://github.com/ayziksha/DSIN

Paper: https://arxiv.org/abs/2001.04753v1
SMOTE Oversampling for Imbalanced Classification with Python

https://machinelearningmastery.com/smote-oversampling-for-imbalanced-classification/
Optuna: A hyperparameter optimization framework

Optuna is an automatic hyperparameter optimization software framework, particularly designed for machine learning.

Code: https://github.com/optuna/optuna

Paper: https://arxiv.org/abs/1907.10902v1

Tutorial: https://optuna.org/
Neural Arithmetic Units

Code for Neural Arithmetic Units (ICLR) and Measuring Arithmetic Extrapolation Performance (SEDL|NeurIPS):

https://github.com/AndreasMadsen/stable-nalu

Paper : https://openreview.net/forum?id=H1gNOeHKPS

@ai_machinelearning_big_data
Plato Dialogue System: A Flexible Conversational AI Research Platform

Code: https://github.com/uber-research/plato-research-dialogue-system

Introducing the Plato Research : https://eng.uber.com/plato-research-dialogue-system/

Paper: https://arxiv.org/abs/2001.06463v1
AI Habitat state-of-the-art simulation platform adds object interactivity
A major update to Facebook AI’s open source AI Habitat platform , which enables significantly faster training of embodied AI agents in a variety of photorealistic 3D virtual environments.

https://ai.facebook.com/blog/ai-habitat-state-of-the-art-simulation-platform-adds-object-interactivity/

Github: https://github.com/facebookresearch/habitat-sim/

https://github.com/facebookresearch/habitat-api

Paper: Are We Making Real Progress in Simulated Environments? Measuring the Sim2Real Gap in Embodied Visual Navigation
https://arxiv.org/abs/1912.06321
This media is not supported in your browser
VIEW IN TELEGRAM
🧠 Releasing the Drosophila Hemibrain Connectome — The Largest Synapse-Resolution Map of Brain Connectivity

https://ai.googleblog.com/2020/01/releasing-drosophila-hemibrain.html
Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation

https://vllab.ucmerced.edu/ym41608/projects/CrossDomainFewShot/

Code and data: https://github.com/hytseng0509/CrossDomainFewShot

Paper: https://arxiv.org/abs/2001.08735
Channel Pruning via Automatic Structure Search

Code: https://github.com/lmbxmu/ABCPruner

Paper: https://arxiv.org/abs/2001.08565
Multi-task self-supervised learning for Robust Speech Recognition

A PASE model can be used as a speech feature extractor or to pre-train an encoder for our desired end-task

Code: https://github.com/santi-pdp/pase

Paper: https://arxiv.org/abs/2001.09239v1

@ai_machinelearning_big_data