A new SotA for generative modelling — Denoising Diffusion Probabilistic Models

Packed BERT: How to accelerate fine-tuning and inference for NLP tasks with packing

Fine-tune GPT-J: a cost-effective GPT-4 alternative for many NLP tasks

Walkthrough tutorial…

Bert-Large Training on the IPU explained

Graphcore engineers explain how they implemented and…

Fine-tune OpenAI’s Whisper Automatic Speech Recognition (ASR) Model

Hugging Face and Graphcore partner for IPU-optimized Transformers

SOTA Transformers meet SOTA hardware

Flan-T5: sweet results with the smaller, more efficient LLM

Flan-T5 offers outstanding performance for…

Exploiting Tile-Level Parallelism on Graphcore IPUs

Getting started with PyTorch Lightning for the IPU

One line of code is all it takes

Why I’m joining Graphcore to lead our Developer Relations

Helping the AI developer community succeed with IPUs