mryab

petals

🌸 Run LLMs at home, BitTorrent-style. Fine-tuning and inference up to 10x faster than offloading

FlexGen

Running large language models on a single GPU for throughput-oriented scenarios.

hivemind

Decentralized deep learning in PyTorch. Built to train models on thousands of volunteers across the world.

cian-datafest2019

2nd place solution for cian.ru Data Fest 2019 competition

collaborative-training

dl-hse-ami

Deep Learning course materials (HSE, Faculty of Computer Science)

efficient-dl-systems

Efficient Deep Learning Systems course materials (HSE, YSDA)

learning-at-home

"Towards Crowdsourced Training of Large Neural Networks using Decentralized Mixture-of-Experts" (NeurIPS 2020), original PyTorch implementation

vmf_loss

PyTorch Implementation of "Von Mises-Fisher Loss for Training Sequence to Sequence Models with Continuous Outputs"

DeDLOC

Official code for "Distributed Deep Learning in Open Collaborations" (NeurIPS 2021)

mryab/efficient-dl-systems

Efficient Deep Learning Systems course materials (HSE, YSDA)

mryab/learning-at-home

"Towards Crowdsourced Training of Large Neural Networks using Decentralized Mixture-of-Experts" (NeurIPS 2020), original PyTorch implementation

mryab/dl-hse-ami

Deep Learning course materials (HSE, Faculty of Computer Science)

mryab/vmf_loss

PyTorch Implementation of "Von Mises-Fisher Loss for Training Sequence to Sequence Models with Continuous Outputs"

mryab/collaborative-training

mryab/cian-datafest2019

2nd place solution for cian.ru Data Fest 2019 competition

mryab/hivemind

Decentralized deep learning in PyTorch. Built to train models on thousands of volunteers across the world.

mryab/mryab.github.io

mryab/transformers

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.